{"doc_key": "ai-dev-1", "ner": [[3, 3, "metrics"], [8, 10, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[8, 10, 3, 3, "part-of", "", true, false]], "relations_mapping_to_source": [0], "sentence": ["Ici", ",", "la", "pr\u00e9cision", "est", "mesur\u00e9e", "par", "le", "taux", "d'", "erreur", ",", "qui", "est", "d\u00e9fini", "comme", "suit", ":"], "sentence-detokenized": "Ici, la pr\u00e9cision est mesur\u00e9e par le taux d'erreur, qui est d\u00e9fini comme suit :", "token2charspan": [[0, 3], [3, 4], [5, 7], [8, 17], [18, 21], [22, 29], [30, 33], [34, 36], [37, 41], [42, 44], [44, 50], [50, 51], [52, 55], [56, 59], [60, 66], [67, 72], [73, 77], [78, 79]]}
{"doc_key": "ai-dev-2", "ner": [[7, 7, "algorithm"], [14, 16, "misc"], [23, 27, "algorithm"], [21, 22, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[7, 7, 14, 16, "type-of", "", false, false], [7, 7, 23, 27, "related-to", "", false, false], [7, 7, 21, 22, "related-to", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["De", "ce", "point", "de", "vue", ",", "le", "SVM", "est", "\u00e9troitement", "li\u00e9", "\u00e0", "d'", "autres", "algorithmes", "de", "classification", "fondamentaux", "tels", "que", "la", "r\u00e9gression", "logistique", "r\u00e9gularis\u00e9e", "par", "les", "moindres", "carr\u00e9s", "."], "sentence-detokenized": "De ce point de vue, le SVM est \u00e9troitement li\u00e9 \u00e0 d'autres algorithmes de classification fondamentaux tels que la r\u00e9gression logistique r\u00e9gularis\u00e9e par les moindres carr\u00e9s.", "token2charspan": [[0, 2], [3, 5], [6, 11], [12, 14], [15, 18], [18, 19], [20, 22], [23, 26], [27, 30], [31, 42], [43, 46], [47, 48], [49, 51], [51, 57], [58, 69], [70, 72], [73, 87], [88, 100], [101, 105], [106, 109], [110, 112], [113, 123], [124, 134], [135, 146], [147, 150], [151, 154], [155, 163], [164, 170], [170, 171]]}
{"doc_key": "ai-dev-3", "ner": [[0, 1, "person"], [3, 4, "person"], [13, 14, "person"], [16, 16, "person"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[3, 4, 0, 1, "named", "actor_plays_character", false, false], [3, 4, 0, 1, "origin", "actor_plays_character", false, false], [16, 16, 13, 14, "named", "actor_plays_character", false, false], [16, 16, 13, 14, "origin", "actor_plays_character", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Brion", "James", "incarne", "Leon", "Kowalski", ",", "un", "r\u00e9plicant", "combattant", "et", "ouvrier", ",", "et", "Joanna", "Cassidy", "incarne", "Zhora", ",", "un", "r\u00e9plicant", "assassin", "."], "sentence-detokenized": "Brion James incarne Leon Kowalski, un r\u00e9plicant combattant et ouvrier, et Joanna Cassidy incarne Zhora, un r\u00e9plicant assassin.", "token2charspan": [[0, 5], [6, 11], [12, 19], [20, 24], [25, 33], [33, 34], [35, 37], [38, 47], [48, 58], [59, 61], [62, 69], [69, 70], [71, 73], [74, 80], [81, 88], [89, 96], [97, 102], [102, 103], [104, 106], [107, 116], [117, 125], [125, 126]]}
{"doc_key": "ai-dev-4", "ner": [[18, 21, "product"], [23, 23, "product"], [26, 26, "organisation"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[18, 21, 26, 26, "physical", "", false, false], [23, 23, 18, 21, "named", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["La", "premi\u00e8re", "image", "\u00e0", "\u00eatre", "scann\u00e9e", ",", "stock\u00e9e", "et", "recr\u00e9\u00e9e", "en", "pixels", "num\u00e9riques", "a", "\u00e9t\u00e9", "affich\u00e9e", "sur", "le", "Standards", "Eastern", "Automatic", "Computer", "(", "SEAC", ")", "au", "NIST", "."], "sentence-detokenized": "La premi\u00e8re image \u00e0 \u00eatre scann\u00e9e, stock\u00e9e et recr\u00e9\u00e9e en pixels num\u00e9riques a \u00e9t\u00e9 affich\u00e9e sur le Standards Eastern Automatic Computer (SEAC) au NIST.", "token2charspan": [[0, 2], [3, 11], [12, 17], [18, 19], [20, 24], [25, 32], [32, 33], [34, 41], [42, 44], [45, 52], [53, 55], [56, 62], [63, 73], [74, 75], [76, 79], [80, 88], [89, 92], [93, 95], [96, 105], [106, 113], [114, 123], [124, 132], [133, 134], [134, 138], [138, 139], [140, 142], [143, 147], [147, 148]]}
{"doc_key": "ai-dev-5", "ner": [[0, 10, "task"], [26, 28, "task"], [31, 32, "task"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[0, 10, 26, 28, "part-of", "", false, false], [0, 10, 31, 32, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["La", "segmentation", "du", "texte", "en", "th\u00e8mes", "ou", "en", "tournures", "de", "discours", "peut", "\u00eatre", "utile", "dans", "certaines", "t\u00e2ches", "de", "traitement", "naturel", ":", "elle", "peut", "am\u00e9liorer", "consid\u00e9rablement", "la", "recherche", "d'", "informations", "ou", "la", "reconnaissance", "vocale", "(", "en", "indexant", "/", "reconnaissant", "les", "documents", "de", "mani\u00e8re", "plus", "pr\u00e9cise", "ou", "en", "donnant", "comme", "r\u00e9sultat", "la", "partie", "sp\u00e9cifique", "d'", "un", "document", "correspondant", "\u00e0", "la", "requ\u00eate", ")", "."], "sentence-detokenized": "La segmentation du texte en th\u00e8mes ou en tournures de discours peut \u00eatre utile dans certaines t\u00e2ches de traitement naturel : elle peut am\u00e9liorer consid\u00e9rablement la recherche d'informations ou la reconnaissance vocale (en indexant/reconnaissant les documents de mani\u00e8re plus pr\u00e9cise ou en donnant comme r\u00e9sultat la partie sp\u00e9cifique d'un document correspondant \u00e0 la requ\u00eate).", "token2charspan": [[0, 2], [3, 15], [16, 18], [19, 24], [25, 27], [28, 34], [35, 37], [38, 40], [41, 50], [51, 53], [54, 62], [63, 67], [68, 72], [73, 78], [79, 83], [84, 93], [94, 100], [101, 103], [104, 114], [115, 122], [123, 124], [125, 129], [130, 134], [135, 144], [145, 161], [162, 164], [165, 174], [175, 177], [177, 189], [190, 192], [193, 195], [196, 210], [211, 217], [218, 219], [219, 221], [222, 230], [230, 231], [231, 244], [245, 248], [249, 258], [259, 261], [262, 269], [270, 274], [275, 282], [283, 285], [286, 288], [289, 296], [297, 302], [303, 311], [312, 314], [315, 321], [322, 332], [333, 335], [335, 337], [338, 346], [347, 360], [361, 362], [363, 365], [366, 373], [373, 374], [374, 375]]}
{"doc_key": "ai-dev-6", "ner": [[11, 13, "university"], [29, 30, "conference"], [34, 36, "university"], [47, 48, "researcher"], [50, 51, "researcher"], [53, 54, "researcher"], [56, 57, "researcher"], [59, 60, "researcher"], [62, 63, "researcher"], [65, 67, "researcher"], [69, 70, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10], "relations": [[29, 30, 34, 36, "physical", "", false, false], [47, 48, 29, 30, "physical", "", false, false], [47, 48, 29, 30, "role", "", false, false], [47, 48, 29, 30, "temporal", "", false, false], [50, 51, 29, 30, "physical", "", false, false], [50, 51, 29, 30, "role", "", false, false], [50, 51, 29, 30, "temporal", "", false, false], [53, 54, 29, 30, "physical", "", false, false], [53, 54, 29, 30, "role", "", false, false], [53, 54, 29, 30, "temporal", "", false, false], [56, 57, 29, 30, "physical", "", false, false], [56, 57, 29, 30, "role", "", false, false], [56, 57, 29, 30, "temporal", "", false, false], [59, 60, 29, 30, "physical", "", false, false], [59, 60, 29, 30, "role", "", false, false], [59, 60, 29, 30, "temporal", "", false, false], [62, 63, 29, 30, "physical", "", false, false], [62, 63, 29, 30, "role", "", false, false], [62, 63, 29, 30, "temporal", "", false, false], [65, 67, 29, 30, "physical", "", false, false], [65, 67, 29, 30, "role", "", false, false], [65, 67, 29, 30, "temporal", "", false, false], [69, 70, 29, 30, "physical", "", false, false], [69, 70, 29, 30, "role", "", false, false], [69, 70, 29, 30, "temporal", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24], "sentence": ["En", "1999", ",", "il", "a", "organis\u00e9", "un", "tel", "symposium", "\u00e0", "l'", "universit\u00e9", "d'", "Indiana", "et", ",", "en", "avril", "2000", ",", "il", "a", "organis\u00e9", "un", "symposium", "plus", "important", "intitul\u00e9", "\"", "Spiritual", "Robots", "\"", "\u00e0", "l'", "universit\u00e9", "de", "Stanford", ",", "dans", "lequel", "il", "a", "anim\u00e9", "un", "panel", "compos\u00e9", "de", "Ray", "Kurzweil", ",", "Hans", "Moravec", ",", "Kevin", "Kelly", ",", "Ralph", "Merkle", ",", "Bill", "Joy", ",", "Frank", "Drake", ",", "John", "Henry", "Holland", "et", "John", "Koza", "."], "sentence-detokenized": "En 1999, il a organis\u00e9 un tel symposium \u00e0 l'universit\u00e9 d'Indiana et, en avril 2000, il a organis\u00e9 un symposium plus important intitul\u00e9 \"Spiritual Robots\" \u00e0 l'universit\u00e9 de Stanford, dans lequel il a anim\u00e9 un panel compos\u00e9 de Ray Kurzweil, Hans Moravec, Kevin Kelly, Ralph Merkle, Bill Joy, Frank Drake, John Henry Holland et John Koza.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 11], [12, 13], [14, 22], [23, 25], [26, 29], [30, 39], [40, 41], [42, 44], [44, 54], [55, 57], [57, 64], [65, 67], [67, 68], [69, 71], [72, 77], [78, 82], [82, 83], [84, 86], [87, 88], [89, 97], [98, 100], [101, 110], [111, 115], [116, 125], [126, 134], [135, 136], [136, 145], [146, 152], [152, 153], [154, 155], [156, 158], [158, 168], [169, 171], [172, 180], [180, 181], [182, 186], [187, 193], [194, 196], [197, 198], [199, 204], [205, 207], [208, 213], [214, 221], [222, 224], [225, 228], [229, 237], [237, 238], [239, 243], [244, 251], [251, 252], [253, 258], [259, 264], [264, 265], [266, 271], [272, 278], [278, 279], [280, 284], [285, 288], [288, 289], [290, 295], [296, 301], [301, 302], [303, 307], [308, 313], [314, 321], [322, 324], [325, 329], [330, 334], [334, 335]]}
{"doc_key": "ai-dev-7", "ner": [[6, 6, "metrics"], [7, 7, "metrics"], [10, 10, "metrics"], [11, 11, "metrics"], [19, 19, "metrics"], [42, 42, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[6, 6, 19, 19, "named", "", false, false], [7, 7, 6, 6, "named", "", false, false], [10, 10, 42, 42, "named", "", false, false], [11, 11, 10, 10, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Il", "consid\u00e8re", "\u00e0", "la", "fois", "la", "pr\u00e9cision", "p", "et", "le", "rappel", "r", "du", "test", "pour", "calculer", "le", "score", ":", "p", "est", "le", "nombre", "de", "r\u00e9sultats", "positifs", "corrects", "divis\u00e9", "par", "le", "nombre", "de", "tous", "les", "r\u00e9sultats", "positifs", "renvoy\u00e9s", "par", "le", "classificateur", ",", "et", "r", "est", "le", "nombre", "de", "r\u00e9sultats", "positifs", "corrects", "divis\u00e9", "par", "le", "nombre", "de", "tous", "les", "\u00e9chantillons", "pertinents", "(", "tous", "les", "\u00e9chantillons", "qui", "auraient", "d\u00fb", "\u00eatre", "identifi\u00e9s", "comme", "positifs", ")", "."], "sentence-detokenized": "Il consid\u00e8re \u00e0 la fois la pr\u00e9cision p et le rappel r du test pour calculer le score : p est le nombre de r\u00e9sultats positifs corrects divis\u00e9 par le nombre de tous les r\u00e9sultats positifs renvoy\u00e9s par le classificateur, et r est le nombre de r\u00e9sultats positifs corrects divis\u00e9 par le nombre de tous les \u00e9chantillons pertinents (tous les \u00e9chantillons qui auraient d\u00fb \u00eatre identifi\u00e9s comme positifs).", "token2charspan": [[0, 2], [3, 12], [13, 14], [15, 17], [18, 22], [23, 25], [26, 35], [36, 37], [38, 40], [41, 43], [44, 50], [51, 52], [53, 55], [56, 60], [61, 65], [66, 74], [75, 77], [78, 83], [84, 85], [86, 87], [88, 91], [92, 94], [95, 101], [102, 104], [105, 114], [115, 123], [124, 132], [133, 139], [140, 143], [144, 146], [147, 153], [154, 156], [157, 161], [162, 165], [166, 175], [176, 184], [185, 193], [194, 197], [198, 200], [201, 215], [215, 216], [217, 219], [220, 221], [222, 225], [226, 228], [229, 235], [236, 238], [239, 248], [249, 257], [258, 266], [267, 273], [274, 277], [278, 280], [281, 287], [288, 290], [291, 295], [296, 299], [300, 312], [313, 323], [324, 325], [325, 329], [330, 333], [334, 346], [347, 350], [351, 359], [360, 362], [363, 367], [368, 378], [379, 384], [385, 393], [393, 394], [394, 395]]}
{"doc_key": "ai-dev-8", "ner": [[4, 4, "organisation"], [26, 26, "product"], [37, 38, "person"], [43, 43, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[4, 4, 26, 26, "artifact", "", false, false], [26, 26, 37, 38, "win-defeat", "", false, false], [26, 26, 43, 43, "win-defeat", "", true, false], [37, 38, 43, 43, "win-defeat", "lose", true, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Depuis", "l'", "acquisition", "de", "Google", ",", "l'", "entreprise", "a", "enregistr\u00e9", "un", "certain", "nombre", "de", "r\u00e9alisations", "importantes", ",", "la", "plus", "notable", "\u00e9tant", "sans", "doute", "la", "cr\u00e9ation", "d'", "AlphaGo", ",", "un", "programme", "qui", "a", "battu", "le", "champion", "du", "monde", "Lee", "Sedol", "au", "jeu", "complexe", "du", "Go", "."], "sentence-detokenized": "Depuis l'acquisition de Google, l'entreprise a enregistr\u00e9 un certain nombre de r\u00e9alisations importantes, la plus notable \u00e9tant sans doute la cr\u00e9ation d'AlphaGo, un programme qui a battu le champion du monde Lee Sedol au jeu complexe du Go.", "token2charspan": [[0, 6], [7, 9], [9, 20], [21, 23], [24, 30], [30, 31], [32, 34], [34, 44], [45, 46], [47, 57], [58, 60], [61, 68], [69, 75], [76, 78], [79, 91], [92, 103], [103, 104], [105, 107], [108, 112], [113, 120], [121, 126], [127, 131], [132, 137], [138, 140], [141, 149], [150, 152], [152, 159], [159, 160], [161, 163], [164, 173], [174, 177], [178, 179], [180, 185], [186, 188], [189, 197], [198, 200], [201, 206], [207, 210], [211, 216], [217, 219], [220, 223], [224, 232], [233, 235], [236, 238], [238, 239]]}
{"doc_key": "ai-dev-9", "ner": [[18, 20, "misc"], [35, 35, "field"], [38, 42, "product"], [60, 60, "misc"], [69, 71, "misc"], [74, 74, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[18, 20, 35, 35, "part-of", "", false, false], [18, 20, 69, 71, "named", "same", false, false], [38, 42, 60, 60, "related-to", "", false, false], [38, 42, 69, 71, "usage", "", false, false], [38, 42, 74, 74, "usage", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["La", "repr\u00e9sentation", "des", "mots", "en", "tenant", "compte", "de", "leur", "contexte", "par", "des", "vecteurs", "denses", "de", "taille", "fixe", "(", "int\u00e9gration", "de", "mots", ")", "est", "devenue", "l'", "un", "des", "blocs", "les", "plus", "fondamentaux", "de", "plusieurs", "syst\u00e8mes", "de", "TAL", ".", "Un", "syst\u00e8me", "de", "d\u00e9sambigu\u00efsation", "non", "supervis\u00e9", "utilise", "la", "similarit\u00e9", "entre", "les", "sens", "des", "mots", "dans", "une", "fen\u00eatre", "de", "contexte", "fixe", "pour", "s\u00e9lectionner", "le", "sens", "le", "plus", "appropri\u00e9", "en", "utilisant", "un", "mod\u00e8le", "d'", "int\u00e9gration", "de", "mots", "pr\u00e9-entra\u00een\u00e9", "et", "WordNet", "."], "sentence-detokenized": "La repr\u00e9sentation des mots en tenant compte de leur contexte par des vecteurs denses de taille fixe (int\u00e9gration de mots) est devenue l'un des blocs les plus fondamentaux de plusieurs syst\u00e8mes de TAL. Un syst\u00e8me de d\u00e9sambigu\u00efsation non supervis\u00e9 utilise la similarit\u00e9 entre les sens des mots dans une fen\u00eatre de contexte fixe pour s\u00e9lectionner le sens le plus appropri\u00e9 en utilisant un mod\u00e8le d'int\u00e9gration de mots pr\u00e9-entra\u00een\u00e9 et WordNet.", "token2charspan": [[0, 2], [3, 17], [18, 21], [22, 26], [27, 29], [30, 36], [37, 43], [44, 46], [47, 51], [52, 60], [61, 64], [65, 68], [69, 77], [78, 84], [85, 87], [88, 94], [95, 99], [100, 101], [101, 112], [113, 115], [116, 120], [120, 121], [122, 125], [126, 133], [134, 136], [136, 138], [139, 142], [143, 148], [149, 152], [153, 157], [158, 170], [171, 173], [174, 183], [184, 192], [193, 195], [196, 199], [199, 200], [201, 203], [204, 211], [212, 214], [215, 231], [232, 235], [236, 245], [246, 253], [254, 256], [257, 267], [268, 273], [274, 277], [278, 282], [283, 286], [287, 291], [292, 296], [297, 300], [301, 308], [309, 311], [312, 320], [321, 325], [326, 330], [331, 343], [344, 346], [347, 351], [352, 354], [355, 359], [360, 369], [370, 372], [373, 382], [383, 385], [386, 392], [393, 395], [395, 406], [407, 409], [410, 414], [415, 427], [428, 430], [431, 438], [438, 439]]}
{"doc_key": "ai-dev-10", "ner": [[3, 4, "field"], [11, 12, "field"], [15, 17, "field"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[11, 12, 3, 4, "part-of", "", false, false], [15, 17, 3, 4, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Les", "techniques", "d'", "apprentissage", "automatique", ",", "qu'", "il", "s'", "agisse", "d'", "apprentissage", "supervis\u00e9", "ou", "d'", "apprentissage", "non", "supervis\u00e9", ",", "ont", "\u00e9t\u00e9", "utilis\u00e9es", "pour", "induire", "ces", "r\u00e8gles", "automatiquement", "."], "sentence-detokenized": "Les techniques d'apprentissage automatique, qu'il s'agisse d'apprentissage supervis\u00e9 ou d'apprentissage non supervis\u00e9, ont \u00e9t\u00e9 utilis\u00e9es pour induire ces r\u00e8gles automatiquement.", "token2charspan": [[0, 3], [4, 14], [15, 17], [17, 30], [31, 42], [42, 43], [44, 47], [47, 49], [50, 52], [52, 58], [59, 61], [61, 74], [75, 84], [85, 87], [88, 90], [90, 103], [104, 107], [108, 117], [117, 118], [119, 122], [123, 126], [127, 136], [137, 141], [142, 149], [150, 153], [154, 160], [161, 176], [176, 177]]}
{"doc_key": "ai-dev-11", "ner": [[3, 4, "researcher"], [7, 9, "product"]], "ner_mapping_to_source": [0, 1], "relations": [[7, 9, 3, 4, "artifact", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["En", "1969", ",", "Scheinman", "a", "invent\u00e9", "le", "bras", "de", "Stanford", ","], "sentence-detokenized": "En 1969, Scheinman a invent\u00e9 le bras de Stanford,", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 18], [19, 20], [21, 28], [29, 31], [32, 36], [37, 39], [40, 48], [48, 49]]}
{"doc_key": "ai-dev-12", "ner": [[2, 3, "metrics"], [8, 12, "misc"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Comme", "la", "perte", "Log", "est", "diff\u00e9rentiable", ",", "une", "m\u00e9thode", "bas\u00e9e", "sur", "le", "gradient", "peut", "\u00eatre", "utilis\u00e9e", "pour", "optimiser", "le", "mod\u00e8le", "."], "sentence-detokenized": "Comme la perte Log est diff\u00e9rentiable, une m\u00e9thode bas\u00e9e sur le gradient peut \u00eatre utilis\u00e9e pour optimiser le mod\u00e8le.", "token2charspan": [[0, 5], [6, 8], [9, 14], [15, 18], [19, 22], [23, 37], [37, 38], [39, 42], [43, 50], [51, 56], [57, 60], [61, 63], [64, 72], [73, 77], [78, 82], [83, 91], [92, 96], [97, 106], [107, 109], [110, 116], [116, 117]]}
{"doc_key": "ai-dev-13", "ner": [[5, 6, "field"], [9, 10, "algorithm"], [12, 12, "algorithm"], [16, 17, "algorithm"], [23, 24, "field"], [37, 37, "task"], [40, 42, "task"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[9, 10, 23, 24, "part-of", "", false, false], [12, 12, 9, 10, "named", "", false, false], [16, 17, 9, 10, "named", "", false, false], [23, 24, 5, 6, "part-of", "subfield", false, false], [37, 37, 23, 24, "part-of", "", false, false], [40, 42, 23, 24, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["Dans", "le", "domaine", "de", "l'", "apprentissage", "automatique", ",", "les", "machines", "support-vecteur", "(", "SVM", ",", "\u00e9galement", "appel\u00e9es", "r\u00e9seaux", "support-vecteur", ")", "sont", "des", "mod\u00e8les", "d'", "apprentissage", "supervis\u00e9", "dot\u00e9s", "d'", "algorithmes", "d'", "apprentissage", "qui", "analysent", "les", "donn\u00e9es", "utilis\u00e9es", "pour", "la", "classification", "et", "l'", "analyse", "de", "r\u00e9gression", "."], "sentence-detokenized": "Dans le domaine de l'apprentissage automatique, les machines support-vecteur (SVM, \u00e9galement appel\u00e9es r\u00e9seaux support-vecteur) sont des mod\u00e8les d'apprentissage supervis\u00e9 dot\u00e9s d'algorithmes d'apprentissage qui analysent les donn\u00e9es utilis\u00e9es pour la classification et l'analyse de r\u00e9gression.", "token2charspan": [[0, 4], [5, 7], [8, 15], [16, 18], [19, 21], [21, 34], [35, 46], [46, 47], [48, 51], [52, 60], [61, 76], [77, 78], [78, 81], [81, 82], [83, 92], [93, 101], [102, 109], [110, 125], [125, 126], [127, 131], [132, 135], [136, 143], [144, 146], [146, 159], [160, 169], [170, 175], [176, 178], [178, 189], [190, 192], [192, 205], [206, 209], [210, 219], [220, 223], [224, 231], [232, 241], [242, 246], [247, 249], [250, 264], [265, 267], [268, 270], [270, 277], [278, 280], [281, 291], [291, 292]]}
{"doc_key": "ai-dev-14", "ner": [[11, 12, "task"], [14, 14, "task"], [32, 32, "metrics"], [34, 34, "metrics"], [36, 36, "researcher"], [38, 38, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[14, 14, 11, 12, "named", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["(", "2002", ")", "comme", "m\u00e9trique", "automatique", "pour", "l'", "\u00e9valuation", "de", "la", "traduction", "automatique", "(", "TA", ")", ",", "de", "nombreuses", "autres", "m\u00e9thodes", "ont", "\u00e9t\u00e9", "propos\u00e9es", "pour", "la", "r\u00e9viser", "ou", "l'", "am\u00e9liorer", ",", "comme", "TER", ",", "METEOR", ",", "Banerjee", "et", "Lavie", "(", "2005", ")", ",", "etc."], "sentence-detokenized": "(2002) comme m\u00e9trique automatique pour l'\u00e9valuation de la traduction automatique (TA), de nombreuses autres m\u00e9thodes ont \u00e9t\u00e9 propos\u00e9es pour la r\u00e9viser ou l'am\u00e9liorer, comme TER, METEOR, Banerjee et Lavie (2005), etc.", "token2charspan": [[0, 1], [1, 5], [5, 6], [7, 12], [13, 21], [22, 33], [34, 38], [39, 41], [41, 51], [52, 54], [55, 57], [58, 68], [69, 80], [81, 82], [82, 84], [84, 85], [85, 86], [87, 89], [90, 100], [101, 107], [108, 116], [117, 120], [121, 124], [125, 134], [135, 139], [140, 142], [143, 150], [151, 153], [154, 156], [156, 165], [165, 166], [167, 172], [173, 176], [176, 177], [178, 184], [184, 185], [186, 194], [195, 197], [198, 203], [204, 205], [205, 209], [209, 210], [210, 211], [212, 216]]}
{"doc_key": "ai-dev-15", "ner": [[3, 4, "misc"], [14, 15, "organisation"], [12, 13, "organisation"], [19, 20, "researcher"], [22, 23, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[3, 4, 12, 13, "origin", "", false, false], [12, 13, 14, 15, "part-of", "", false, false], [19, 20, 12, 13, "role", "", false, false], [22, 23, 12, 13, "role", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Il", "comprend", "une", "ontologie", "sup\u00e9rieure", ",", "cr\u00e9\u00e9e", "par", "le", "groupe", "de", "travail", "P1600.1", "de", "l'", "IEEE", "(", "initialement", "par", "Ian", "Niles", "et", "Adam", "Pease", ")", "."], "sentence-detokenized": "Il comprend une ontologie sup\u00e9rieure, cr\u00e9\u00e9e par le groupe de travail P1600.1 de l'IEEE (initialement par Ian Niles et Adam Pease).", "token2charspan": [[0, 2], [3, 11], [12, 15], [16, 25], [26, 36], [36, 37], [38, 43], [44, 47], [48, 50], [51, 57], [58, 60], [61, 68], [69, 76], [77, 79], [80, 82], [82, 86], [87, 88], [88, 100], [101, 104], [105, 108], [109, 114], [115, 117], [118, 122], [123, 128], [128, 129], [129, 130]]}
{"doc_key": "ai-dev-16", "ner": [[1, 2, "misc"], [32, 35, "algorithm"], [38, 40, "algorithm"], [46, 48, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[32, 35, 1, 2, "part-of", "", true, false], [38, 40, 1, 2, "part-of", "", true, false], [46, 48, 38, 40, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["En", "cryo-tomographie", "\u00e9lectronique", ",", "o\u00f9", "un", "nombre", "limit\u00e9", "de", "projections", "est", "acquis", "en", "raison", "des", "limitations", "mat\u00e9rielles", "et", "pour", "\u00e9viter", "d'", "endommager", "les", "sp\u00e9cimens", "biologiques", ",", "elle", "peut", "\u00eatre", "utilis\u00e9e", "avec", "des", "techniques", "de", "d\u00e9tection", "compressive", "ou", "des", "fonctions", "de", "r\u00e9gularisation", "(", "par", "exemple", ",", "la", "perte", "de", "Huber", ")", "pour", "am\u00e9liorer", "la", "reconstruction", "et", "permettre", "une", "meilleure", "interpr\u00e9tation", "."], "sentence-detokenized": "En cryo-tomographie \u00e9lectronique, o\u00f9 un nombre limit\u00e9 de projections est acquis en raison des limitations mat\u00e9rielles et pour \u00e9viter d'endommager les sp\u00e9cimens biologiques, elle peut \u00eatre utilis\u00e9e avec des techniques de d\u00e9tection compressive ou des fonctions de r\u00e9gularisation (par exemple, la perte de Huber) pour am\u00e9liorer la reconstruction et permettre une meilleure interpr\u00e9tation.", "token2charspan": [[0, 2], [3, 19], [20, 32], [32, 33], [34, 36], [37, 39], [40, 46], [47, 53], [54, 56], [57, 68], [69, 72], [73, 79], [80, 82], [83, 89], [90, 93], [94, 105], [106, 117], [118, 120], [121, 125], [126, 132], [133, 135], [135, 145], [146, 149], [150, 159], [160, 171], [171, 172], [173, 177], [178, 182], [183, 187], [188, 196], [197, 201], [202, 205], [206, 216], [217, 219], [220, 229], [230, 241], [242, 244], [245, 248], [249, 258], [259, 261], [262, 276], [277, 278], [278, 281], [282, 289], [289, 290], [291, 293], [294, 299], [300, 302], [303, 308], [308, 309], [310, 314], [315, 324], [325, 327], [328, 342], [343, 345], [346, 355], [356, 359], [360, 369], [370, 384], [384, 385]]}
{"doc_key": "ai-dev-17", "ner": [[6, 6, "misc"], [8, 8, "programlang"], [13, 14, "algorithm"], [17, 18, "algorithm"], [22, 23, "algorithm"], [29, 31, "product"], [34, 34, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[6, 6, 8, 8, "part-of", "", false, false], [13, 14, 6, 6, "type-of", "", false, false], [17, 18, 6, 6, "type-of", "", false, false], [22, 23, 6, 6, "type-of", "", false, false], [29, 31, 8, 8, "general-affiliation", "", true, false], [29, 31, 8, 8, "part-of", "", true, false], [34, 34, 29, 31, "role", "publishes", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "sentence": ["Une", "impl\u00e9mentation", "de", "plusieurs", "proc\u00e9dures", "de", "blanchiment", "dans", "R", ",", "y", "compris", "le", "blanchiment", "ZCA", "et", "le", "blanchiment", "PCA", "mais", "aussi", "le", "blanchiment", "CCA", ",", "est", "disponible", "dans", "le", "paquet", "R", "whitening", "publi\u00e9", "sur", "CRAN", "."], "sentence-detokenized": "Une impl\u00e9mentation de plusieurs proc\u00e9dures de blanchiment dans R, y compris le blanchiment ZCA et le blanchiment PCA mais aussi le blanchiment CCA, est disponible dans le paquet R whitening publi\u00e9 sur CRAN.", "token2charspan": [[0, 3], [4, 18], [19, 21], [22, 31], [32, 42], [43, 45], [46, 57], [58, 62], [63, 64], [64, 65], [66, 67], [68, 75], [76, 78], [79, 90], [91, 94], [95, 97], [98, 100], [101, 112], [113, 116], [117, 121], [122, 127], [128, 130], [131, 142], [143, 146], [146, 147], [148, 151], [152, 162], [163, 167], [168, 170], [171, 177], [178, 179], [180, 189], [190, 196], [197, 200], [201, 205], [205, 206]]}
{"doc_key": "ai-dev-18", "ner": [[34, 34, "product"], [36, 36, "product"], [38, 38, "product"], [40, 40, "product"], [42, 42, "product"], [44, 44, "product"], [48, 49, "programlang"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[34, 34, 38, 38, "compare", "", false, false], [34, 34, 40, 40, "compare", "", false, false], [34, 34, 42, 42, "compare", "", false, false], [34, 34, 44, 44, "compare", "", false, false], [34, 34, 48, 49, "compare", "", false, false], [36, 36, 38, 38, "compare", "", false, false], [36, 36, 40, 40, "compare", "", false, false], [36, 36, 42, 42, "compare", "", false, false], [36, 36, 44, 44, "compare", "", false, false], [36, 36, 48, 49, "compare", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9], "sentence": ["Aujourd'hui", ",", "le", "domaine", "est", "devenu", "encore", "plus", "intimidant", "et", "complexe", "avec", "l'", "ajout", "de", "langages", "et", "de", "logiciels", "d'", "analyse", "et", "de", "conception", "de", "circuits", ",", "de", "syst\u00e8mes", "et", "de", "signaux", ",", "de", "MATLAB", "et", "Simulink", "\u00e0", "NumPy", ",", "VHDL", ",", "PSpice", ",", "Verilog", "et", "m\u00eame", "au", "langage", "assembleur", "."], "sentence-detokenized": "Aujourd'hui, le domaine est devenu encore plus intimidant et complexe avec l'ajout de langages et de logiciels d'analyse et de conception de circuits, de syst\u00e8mes et de signaux, de MATLAB et Simulink \u00e0 NumPy, VHDL, PSpice, Verilog et m\u00eame au langage assembleur.", "token2charspan": [[0, 11], [11, 12], [13, 15], [16, 23], [24, 27], [28, 34], [35, 41], [42, 46], [47, 57], [58, 60], [61, 69], [70, 74], [75, 77], [77, 82], [83, 85], [86, 94], [95, 97], [98, 100], [101, 110], [111, 113], [113, 120], [121, 123], [124, 126], [127, 137], [138, 140], [141, 149], [149, 150], [151, 153], [154, 162], [163, 165], [166, 168], [169, 176], [176, 177], [178, 180], [181, 187], [188, 190], [191, 199], [200, 201], [202, 207], [207, 208], [209, 213], [213, 214], [215, 221], [221, 222], [223, 230], [231, 233], [234, 238], [239, 241], [242, 249], [250, 260], [260, 261]]}
{"doc_key": "ai-dev-19", "ner": [[6, 7, "person"], [21, 22, "person"], [18, 20, "organisation"], [26, 26, "product"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[18, 20, 21, 22, "origin", "", false, false], [26, 26, 18, 20, "artifact", "builds", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["L'", "entreprise", "a", "\u00e9t\u00e9", "fond\u00e9e", "par", "Kiichiro", "Toyoda", "en", "1937", ",", "en", "tant", "que", "spin-off", "de", "la", "soci\u00e9t\u00e9", "Toyota", "Industries", "de", "Sakichi", "Toyoda", "pour", "cr\u00e9er", "des", "automobiles", "."], "sentence-detokenized": "L'entreprise a \u00e9t\u00e9 fond\u00e9e par Kiichiro Toyoda en 1937, en tant que spin-off de la soci\u00e9t\u00e9 Toyota Industries de Sakichi Toyoda pour cr\u00e9er des automobiles.", "token2charspan": [[0, 2], [2, 12], [13, 14], [15, 18], [19, 25], [26, 29], [30, 38], [39, 45], [46, 48], [49, 53], [53, 54], [55, 57], [58, 62], [63, 66], [67, 75], [76, 78], [79, 81], [82, 89], [90, 96], [97, 107], [108, 110], [111, 118], [119, 125], [126, 130], [131, 136], [137, 140], [141, 152], [152, 153]]}
{"doc_key": "ai-dev-20", "ner": [[0, 3, "field"], [62, 63, "field"]], "ner_mapping_to_source": [0, 1], "relations": [[62, 63, 0, 3, "origin", "", true, false]], "relations_mapping_to_source": [0], "sentence": ["L'", "apprentissage", "non", "supervis\u00e9", ",", "quant", "\u00e0", "lui", ",", "se", "base", "sur", "des", "donn\u00e9es", "d'", "apprentissage", "qui", "n'", "ont", "pas", "\u00e9t\u00e9", "\u00e9tiquet\u00e9es", "manuellement", "et", "tente", "de", "trouver", "des", "mod\u00e8les", "inh\u00e9rents", "aux", "donn\u00e9es", "qui", "peuvent", "ensuite", "\u00eatre", "utilis\u00e9s", "pour", "d\u00e9terminer", "la", "valeur", "de", "sortie", "correcte", "pour", "les", "nouvelles", "instances", "de", "donn\u00e9es", ".", "Une", "combinaison", "des", "deux", "qui", "a", "\u00e9t\u00e9", "r\u00e9cemment", "explor\u00e9e", "est", "l'", "apprentissage", "semi-supervis\u00e9", ",", "qui", "utilise", "une", "combinaison", "de", "donn\u00e9es", "\u00e9tiquet\u00e9es", "et", "non", "\u00e9tiquet\u00e9es", "(", "g\u00e9n\u00e9ralement", "un", "petit", "ensemble", "de", "donn\u00e9es", "\u00e9tiquet\u00e9es", "combin\u00e9", "\u00e0", "une", "grande", "quantit\u00e9", "de", "donn\u00e9es", "non", "\u00e9tiquet\u00e9es", ")", "."], "sentence-detokenized": "L'apprentissage non supervis\u00e9, quant \u00e0 lui, se base sur des donn\u00e9es d'apprentissage qui n'ont pas \u00e9t\u00e9 \u00e9tiquet\u00e9es manuellement et tente de trouver des mod\u00e8les inh\u00e9rents aux donn\u00e9es qui peuvent ensuite \u00eatre utilis\u00e9s pour d\u00e9terminer la valeur de sortie correcte pour les nouvelles instances de donn\u00e9es . Une combinaison des deux qui a \u00e9t\u00e9 r\u00e9cemment explor\u00e9e est l'apprentissage semi-supervis\u00e9, qui utilise une combinaison de donn\u00e9es \u00e9tiquet\u00e9es et non \u00e9tiquet\u00e9es (g\u00e9n\u00e9ralement un petit ensemble de donn\u00e9es \u00e9tiquet\u00e9es combin\u00e9 \u00e0 une grande quantit\u00e9 de donn\u00e9es non \u00e9tiquet\u00e9es).", "token2charspan": [[0, 2], [2, 15], [16, 19], [20, 29], [29, 30], [31, 36], [37, 38], [39, 42], [42, 43], [44, 46], [47, 51], [52, 55], [56, 59], [60, 67], [68, 70], [70, 83], [84, 87], [88, 90], [90, 93], [94, 97], [98, 101], [102, 112], [113, 125], [126, 128], [129, 134], [135, 137], [138, 145], [146, 149], [150, 157], [158, 167], [168, 171], [172, 179], [180, 183], [184, 191], [192, 199], [200, 204], [205, 213], [214, 218], [219, 229], [230, 232], [233, 239], [240, 242], [243, 249], [250, 258], [259, 263], [264, 267], [268, 277], [278, 287], [288, 290], [291, 298], [299, 300], [301, 304], [305, 316], [317, 320], [321, 325], [326, 329], [330, 331], [332, 335], [336, 345], [346, 354], [355, 358], [359, 361], [361, 374], [375, 389], [389, 390], [391, 394], [395, 402], [403, 406], [407, 418], [419, 421], [422, 429], [430, 440], [441, 443], [444, 447], [448, 458], [459, 460], [460, 472], [473, 475], [476, 481], [482, 490], [491, 493], [494, 501], [502, 512], [513, 520], [521, 522], [523, 526], [527, 533], [534, 542], [543, 545], [546, 553], [554, 557], [558, 568], [568, 569], [569, 570]]}
{"doc_key": "ai-dev-21", "ner": [[21, 21, "organisation"], [19, 19, "product"], [26, 27, "organisation"], [24, 24, "product"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[19, 19, 21, 21, "artifact", "", false, false], [26, 27, 24, 24, "artifact", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Malgr\u00e9", "ces", "robots", "humano\u00efdes", "\u00e0", "usage", "utilitaire", ",", "il", "existe", "des", "robots", "humano\u00efdes", "destin\u00e9s", "au", "divertissement", ",", "comme", "le", "QRIO", "de", "Sony", "et", "le", "RoboSapien", "de", "Wow", "Wee", "."], "sentence-detokenized": "Malgr\u00e9 ces robots humano\u00efdes \u00e0 usage utilitaire, il existe des robots humano\u00efdes destin\u00e9s au divertissement, comme le QRIO de Sony et le RoboSapien de Wow Wee.", "token2charspan": [[0, 6], [7, 10], [11, 17], [18, 28], [29, 30], [31, 36], [37, 47], [47, 48], [49, 51], [52, 58], [59, 62], [63, 69], [70, 80], [81, 89], [90, 92], [93, 107], [107, 108], [109, 114], [115, 117], [118, 122], [123, 125], [126, 130], [131, 133], [134, 136], [137, 147], [148, 150], [151, 154], [155, 158], [158, 159]]}
{"doc_key": "ai-dev-22", "ner": [[0, 0, "researcher"], [6, 12, "conference"]], "ner_mapping_to_source": [0, 1], "relations": [[0, 0, 6, 12, "role", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Webber", "est", "devenu", "membre", "de", "l'", "Association", "for", "the", "Advancement", "of", "Artificial", "Intelligence", "en", "1991", ","], "sentence-detokenized": "Webber est devenu membre de l'Association for the Advancement of Artificial Intelligence en 1991,", "token2charspan": [[0, 6], [7, 10], [11, 17], [18, 24], [25, 27], [28, 30], [30, 41], [42, 45], [46, 49], [50, 61], [62, 64], [65, 75], [76, 88], [89, 91], [92, 96], [96, 97]]}
{"doc_key": "ai-dev-23", "ner": [[10, 12, "field"], [15, 17, "field"], [31, 35, "task"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[31, 35, 10, 12, "part-of", "task_part_of_field", false, false], [31, 35, 15, 17, "part-of", "task_part_of_field", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Avec", "cette", "entreprise", ",", "il", "a", "d\u00e9velopp\u00e9", "une", "technologie", "d'", "exploration", "de", "donn\u00e9es", "et", "de", "bases", "de", "donn\u00e9es", ",", "plus", "sp\u00e9cifiquement", "des", "ontologies", "de", "haut", "niveau", "pour", "l'", "intelligence", "et", "la", "compr\u00e9hension", "automatique", "du", "langage", "naturel", "."], "sentence-detokenized": "Avec cette entreprise, il a d\u00e9velopp\u00e9 une technologie d'exploration de donn\u00e9es et de bases de donn\u00e9es, plus sp\u00e9cifiquement des ontologies de haut niveau pour l'intelligence et la compr\u00e9hension automatique du langage naturel.", "token2charspan": [[0, 4], [5, 10], [11, 21], [21, 22], [23, 25], [26, 27], [28, 37], [38, 41], [42, 53], [54, 56], [56, 67], [68, 70], [71, 78], [79, 81], [82, 84], [85, 90], [91, 93], [94, 101], [101, 102], [103, 107], [108, 122], [123, 126], [127, 137], [138, 140], [141, 145], [146, 152], [153, 157], [158, 160], [160, 172], [173, 175], [176, 178], [179, 192], [193, 204], [205, 207], [208, 215], [216, 223], [223, 224]]}
{"doc_key": "ai-dev-24", "ner": [[30, 31, "misc"], [34, 37, "misc"], [40, 41, "misc"], [44, 44, "country"], [47, 50, "organisation"], [52, 52, "country"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[30, 31, 44, 44, "physical", "", false, false], [34, 37, 44, 44, "physical", "", false, false], [40, 41, 44, 44, "physical", "", false, false], [47, 50, 52, 52, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Toutefois", ",", "au", "cours", "des", "derni\u00e8res", "ann\u00e9es", ",", "on", "a", "pu", "observer", "l'", "apparition", "de", "diff\u00e9rents", "services", "\u00e9lectroniques", "et", "d'", "initiatives", "connexes", "dans", "les", "pays", "en", "d\u00e9veloppement", ",", "comme", "le", "projet", "Nemmadi", ",", "le", "projet", "MCA21", "Mission", "Mode", "ou", "encore", "Digital", "India", ",", "en", "Inde", ";", "la", "Direction", "du", "gouvernement", "\u00e9lectronique", "au", "Pakistan", ";", "etc."], "sentence-detokenized": "Toutefois, au cours des derni\u00e8res ann\u00e9es, on a pu observer l'apparition de diff\u00e9rents services \u00e9lectroniques et d'initiatives connexes dans les pays en d\u00e9veloppement, comme le projet Nemmadi, le projet MCA21 Mission Mode ou encore Digital India, en Inde ; la Direction du gouvernement \u00e9lectronique au Pakistan ; etc.", "token2charspan": [[0, 9], [9, 10], [11, 13], [14, 19], [20, 23], [24, 33], [34, 40], [40, 41], [42, 44], [45, 46], [47, 49], [50, 58], [59, 61], [61, 71], [72, 74], [75, 85], [86, 94], [95, 108], [109, 111], [112, 114], [114, 125], [126, 134], [135, 139], [140, 143], [144, 148], [149, 151], [152, 165], [165, 166], [167, 172], [173, 175], [176, 182], [183, 190], [190, 191], [192, 194], [195, 201], [202, 207], [208, 215], [216, 220], [221, 223], [224, 230], [231, 238], [239, 244], [244, 245], [246, 248], [249, 253], [254, 255], [256, 258], [259, 268], [269, 271], [272, 284], [285, 297], [298, 300], [301, 309], [310, 311], [312, 316]]}
{"doc_key": "ai-dev-25", "ner": [[4, 4, "misc"], [6, 6, "field"], [8, 8, "field"], [11, 13, "university"], [16, 18, "university"], [27, 29, "university"], [34, 34, "misc"], [36, 37, "field"], [41, 45, "misc"], [46, 46, "university"], [48, 50, "university"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10], "relations": [[4, 4, 6, 6, "topic", "", false, false], [4, 4, 8, 8, "topic", "", false, false], [4, 4, 11, 13, "origin", "", false, false], [11, 13, 16, 18, "part-of", "", false, false], [27, 29, 11, 13, "part-of", "", false, false], [34, 34, 36, 37, "topic", "", false, false], [34, 34, 46, 46, "origin", "", false, false], [41, 45, 46, 46, "origin", "", false, false], [46, 46, 48, 50, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "sentence": ["Il", "a", "obtenu", "un", "doctorat", "en", "radiophysique", "et", "\u00e9lectronique", "du", "campus", "Rajabazar", "Science", "College", "de", "l'", "universit\u00e9", "de", "Calcutta", "en", "1979", "en", "tant", "qu'", "\u00e9tudiant", "de", "l'", "Indian", "Statistical", "Institute", ",", "et", "un", "autre", "doctorat", "en", "g\u00e9nie", "\u00e9lectrique", "ainsi", "qu'", "un", "dipl\u00f4me", "de", "l'", "Imperial", "College", "de", "l'", "universit\u00e9", "de", "Londres", "en", "1982", "."], "sentence-detokenized": "Il a obtenu un doctorat en radiophysique et \u00e9lectronique du campus Rajabazar Science College de l'universit\u00e9 de Calcutta en 1979 en tant qu'\u00e9tudiant de l'Indian Statistical Institute, et un autre doctorat en g\u00e9nie \u00e9lectrique ainsi qu'un dipl\u00f4me de l'Imperial College de l'universit\u00e9 de Londres en 1982.", "token2charspan": [[0, 2], [3, 4], [5, 11], [12, 14], [15, 23], [24, 26], [27, 40], [41, 43], [44, 56], [57, 59], [60, 66], [67, 76], [77, 84], [85, 92], [93, 95], [96, 98], [98, 108], [109, 111], [112, 120], [121, 123], [124, 128], [129, 131], [132, 136], [137, 140], [140, 148], [149, 151], [152, 154], [154, 160], [161, 172], [173, 182], [182, 183], [184, 186], [187, 189], [190, 195], [196, 204], [205, 207], [208, 213], [214, 224], [225, 230], [231, 234], [234, 236], [237, 244], [245, 247], [248, 250], [250, 258], [259, 266], [267, 269], [270, 272], [272, 282], [283, 285], [286, 293], [294, 296], [297, 301], [301, 302]]}
{"doc_key": "ai-dev-26", "ner": [[0, 2, "location"], [23, 25, "misc"], [31, 32, "misc"], [34, 36, "person"], [38, 39, "person"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[23, 25, 0, 2, "temporal", "", false, false], [31, 32, 0, 2, "temporal", "", false, false], [34, 36, 31, 32, "role", "actor_in", false, false], [38, 39, 31, 32, "role", "actor_in", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["L'", "Expo", "II", "a", "\u00e9t\u00e9", "annonc\u00e9e", "comme", "\u00e9tant", "le", "lieu", "de", "la", "premi\u00e8re", "mondiale", "de", "plusieurs", "films", "jamais", "vus", "en", "3D", ",", "dont", "The", "Diamond", "Wizard", "et", "le", "court-m\u00e9trage", "Universal", ",", "Hawaiian", "Nights", "with", "Mamie", "Van", "Doren", "and", "Pinky", "Lee", "."], "sentence-detokenized": "L'Expo II a \u00e9t\u00e9 annonc\u00e9e comme \u00e9tant le lieu de la premi\u00e8re mondiale de plusieurs films jamais vus en 3D, dont The Diamond Wizard et le court-m\u00e9trage Universal, Hawaiian Nights with Mamie Van Doren and Pinky Lee.", "token2charspan": [[0, 2], [2, 6], [7, 9], [10, 11], [12, 15], [16, 24], [25, 30], [31, 36], [37, 39], [40, 44], [45, 47], [48, 50], [51, 59], [60, 68], [69, 71], [72, 81], [82, 87], [88, 94], [95, 98], [99, 101], [102, 104], [104, 105], [106, 110], [111, 114], [115, 122], [123, 129], [130, 132], [133, 135], [136, 149], [150, 159], [159, 160], [161, 169], [170, 176], [177, 181], [182, 187], [188, 191], [192, 197], [198, 201], [202, 207], [208, 211], [211, 212]]}
{"doc_key": "ai-dev-27", "ner": [[10, 11, "researcher"], [20, 24, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[10, 11, 20, 24, "related-to", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Le", "probl\u00e8me", "du", "maximum", "de", "sous-r\u00e9seaux", "a", "\u00e9t\u00e9", "propos\u00e9", "par", "Ulf", "Grenander", "en", "1977", "comme", "un", "mod\u00e8le", "simplifi\u00e9", "pour", "l'", "estimation", "par", "maximum", "de", "vraisemblance", "de", "motifs", "dans", "des", "images", "num\u00e9ris\u00e9es", "."], "sentence-detokenized": "Le probl\u00e8me du maximum de sous-r\u00e9seaux a \u00e9t\u00e9 propos\u00e9 par Ulf Grenander en 1977 comme un mod\u00e8le simplifi\u00e9 pour l'estimation par maximum de vraisemblance de motifs dans des images num\u00e9ris\u00e9es.", "token2charspan": [[0, 2], [3, 11], [12, 14], [15, 22], [23, 25], [26, 38], [39, 40], [41, 44], [45, 52], [53, 56], [57, 60], [61, 70], [71, 73], [74, 78], [79, 84], [85, 87], [88, 94], [95, 104], [105, 109], [110, 112], [112, 122], [123, 126], [127, 134], [135, 137], [138, 151], [152, 154], [155, 161], [162, 166], [167, 170], [171, 177], [178, 188], [188, 189]]}
{"doc_key": "ai-dev-28", "ner": [[1, 2, "product"], [5, 6, "product"], [9, 11, "product"], [14, 15, "product"], [18, 20, "product"], [23, 25, "product"], [40, 40, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[40, 40, 1, 2, "part-of", "", false, false], [40, 40, 5, 6, "part-of", "", false, false], [40, 40, 9, 11, "part-of", "", false, false], [40, 40, 14, 15, "part-of", "", false, false], [40, 40, 18, 20, "part-of", "", false, false], [40, 40, 23, 25, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["L'", "iPhone", "4S", ",", "l'", "iPad", "3", ",", "l'", "iPad", "Mini", "1G", ",", "l'", "iPad", "Air", ",", "l'", "iPad", "Pro", "1G", ",", "l'", "iPod", "Touch", "5G", "et", "les", "versions", "ult\u00e9rieures", "sont", "tous", "\u00e9quip\u00e9s", "d'", "un", "assistant", "vocal", "plus", "avanc\u00e9", "appel\u00e9", "Siri", "."], "sentence-detokenized": "L'iPhone 4S, l'iPad 3, l'iPad Mini 1G, l'iPad Air, l'iPad Pro 1G, l'iPod Touch 5G et les versions ult\u00e9rieures sont tous \u00e9quip\u00e9s d'un assistant vocal plus avanc\u00e9 appel\u00e9 Siri.", "token2charspan": [[0, 2], [2, 8], [9, 11], [11, 12], [13, 15], [15, 19], [20, 21], [21, 22], [23, 25], [25, 29], [30, 34], [35, 37], [37, 38], [39, 41], [41, 45], [46, 49], [49, 50], [51, 53], [53, 57], [58, 61], [62, 64], [64, 65], [66, 68], [68, 72], [73, 78], [79, 81], [82, 84], [85, 88], [89, 97], [98, 109], [110, 114], [115, 119], [120, 127], [128, 130], [130, 132], [133, 142], [143, 148], [149, 153], [154, 160], [161, 167], [168, 172], [172, 173]]}
{"doc_key": "ai-dev-29", "ner": [[7, 8, "metrics"], [11, 15, "metrics"], [17, 18, "metrics"], [45, 48, "metrics"], [54, 56, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[11, 15, 45, 48, "named", "", false, false], [17, 18, 11, 15, "named", "", false, false], [45, 48, 54, 56, "related-to", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Il", "est", "facile", "de", "v\u00e9rifier", "que", "la", "perte", "logistique", "et", "la", "perte", "d'", "entropie", "crois\u00e9e", "binaire", "(", "perte", "Log", ")", "sont", "en", "fait", "les", "m\u00eames", "(", "jusqu'", "\u00e0", "une", "constante", "multiplicative", "math", "\\", "frac", "{1}", "{", "\\", "log", "(", "2)}", "/", "math", ")", ".", "{La", "perte", "d'", "entropie", "crois\u00e9e", "est", "\u00e9troitement", "li\u00e9e", "\u00e0", "la", "divergence", "de", "Kullback-Leibler", "entre", "la", "distribution", "empirique", "et", "la", "distribution", "pr\u00e9dite", "."], "sentence-detokenized": "Il est facile de v\u00e9rifier que la perte logistique et la perte d'entropie crois\u00e9e binaire (perte Log) sont en fait les m\u00eames (jusqu'\u00e0 une constante multiplicative math\\ frac {1} {\\ log (2)} / math). {La perte d'entropie crois\u00e9e est \u00e9troitement li\u00e9e \u00e0 la divergence de Kullback-Leibler entre la distribution empirique et la distribution pr\u00e9dite.", "token2charspan": [[0, 2], [3, 6], [7, 13], [14, 16], [17, 25], [26, 29], [30, 32], [33, 38], [39, 49], [50, 52], [53, 55], [56, 61], [62, 64], [64, 72], [73, 80], [81, 88], [89, 90], [90, 95], [96, 99], [99, 100], [101, 105], [106, 108], [109, 113], [114, 117], [118, 123], [124, 125], [125, 131], [131, 132], [133, 136], [137, 146], [147, 161], [162, 166], [166, 167], [168, 172], [173, 176], [177, 178], [178, 179], [180, 183], [184, 185], [185, 188], [189, 190], [191, 195], [195, 196], [196, 197], [198, 201], [202, 207], [208, 210], [210, 218], [219, 226], [227, 230], [231, 242], [243, 247], [248, 249], [250, 252], [253, 263], [264, 266], [267, 283], [284, 289], [290, 292], [293, 305], [306, 315], [316, 318], [319, 321], [322, 334], [335, 342], [342, 343]]}
{"doc_key": "ai-dev-30", "ner": [[1, 2, "algorithm"], [13, 15, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[13, 15, 1, 2, "usage", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["L'", "algorithme", "EM", "est", "utilis\u00e9", "pour", "trouver", "les", "param\u00e8tres", "(", "locaux", ")", "de", "maximum", "de", "vraisemblance", "d'", "un", "mod\u00e8le", "statistique", "dans", "les", "cas", "o\u00f9", "les", "\u00e9quations", "ne", "peuvent", "\u00eatre", "r\u00e9solues", "directement", "."], "sentence-detokenized": "L'algorithme EM est utilis\u00e9 pour trouver les param\u00e8tres (locaux) de maximum de vraisemblance d'un mod\u00e8le statistique dans les cas o\u00f9 les \u00e9quations ne peuvent \u00eatre r\u00e9solues directement.", "token2charspan": [[0, 2], [2, 12], [13, 15], [16, 19], [20, 27], [28, 32], [33, 40], [41, 44], [45, 55], [56, 57], [57, 63], [63, 64], [65, 67], [68, 75], [76, 78], [79, 92], [93, 95], [95, 97], [98, 104], [105, 116], [117, 121], [122, 125], [126, 129], [130, 132], [133, 136], [137, 146], [147, 149], [150, 157], [158, 162], [163, 171], [172, 183], [183, 184]]}
{"doc_key": "ai-dev-31", "ner": [[12, 15, "task"], [18, 23, "task"], [30, 37, "task"], [34, 36, "task"], [44, 51, "task"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [], "relations_mapping_to_source": [], "sentence": ["Ces", "recherches", "ont", "\u00e9t\u00e9", "fondamentales", "pour", "le", "d\u00e9veloppement", "des", "techniques", "modernes", "de", "synth\u00e8se", "de", "la", "parole", ",", "des", "machines", "\u00e0", "lire", "pour", "les", "aveugles", ",", "de", "l'", "\u00e9tude", "de", "la", "perception", "et", "de", "la", "reconnaissance", "de", "la", "parole", ",", "et", "du", "d\u00e9veloppement", "de", "la", "th\u00e9orie", "motrice", "de", "la", "perception", "de", "la", "parole", "."], "sentence-detokenized": "Ces recherches ont \u00e9t\u00e9 fondamentales pour le d\u00e9veloppement des techniques modernes de synth\u00e8se de la parole, des machines \u00e0 lire pour les aveugles, de l'\u00e9tude de la perception et de la reconnaissance de la parole, et du d\u00e9veloppement de la th\u00e9orie motrice de la perception de la parole.", "token2charspan": [[0, 3], [4, 14], [15, 18], [19, 22], [23, 36], [37, 41], [42, 44], [45, 58], [59, 62], [63, 73], [74, 82], [83, 85], [86, 94], [95, 97], [98, 100], [101, 107], [107, 108], [109, 112], [113, 121], [122, 123], [124, 128], [129, 133], [134, 137], [138, 146], [146, 147], [148, 150], [151, 153], [153, 158], [159, 161], [162, 164], [165, 175], [176, 178], [179, 181], [182, 184], [185, 199], [200, 202], [203, 205], [206, 212], [212, 213], [214, 216], [217, 219], [220, 233], [234, 236], [237, 239], [240, 247], [248, 255], [256, 258], [259, 261], [262, 272], [273, 275], [276, 278], [279, 285], [285, 286]]}
{"doc_key": "ai-dev-32", "ner": [[8, 8, "product"], [1, 4, "misc"], [6, 6, "misc"], [11, 12, "misc"], [15, 15, "product"], [17, 17, "product"], [19, 19, "product"], [27, 27, "programlang"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "relations": [[1, 4, 8, 8, "origin", "", false, false], [1, 4, 11, 12, "type-of", "", false, false], [1, 4, 15, 15, "related-to", "program_for", false, false], [1, 4, 17, 17, "related-to", "program_for", false, false], [1, 4, 19, 19, "related-to", "program_for", false, false], [1, 4, 27, 27, "related-to", "program_for", false, false], [6, 6, 1, 4, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "sentence": ["L'", "environnement", "de", "d\u00e9veloppement", "int\u00e9gr\u00e9", "(", "IDE", ")", "Arduino", "est", "une", "application", "multiplateforme", "(", "pour", "Windows", ",", "macOS", "et", "Linux", ")", "\u00e9crite", "dans", "le", "langage", "de", "programmation", "Java", "."], "sentence-detokenized": "L'environnement de d\u00e9veloppement int\u00e9gr\u00e9 (IDE) Arduino est une application multiplateforme (pour Windows, macOS et Linux) \u00e9crite dans le langage de programmation Java.", "token2charspan": [[0, 2], [2, 15], [16, 18], [19, 32], [33, 40], [41, 42], [42, 45], [45, 46], [47, 54], [55, 58], [59, 62], [63, 74], [75, 90], [91, 92], [92, 96], [97, 104], [104, 105], [106, 111], [112, 114], [115, 120], [120, 121], [122, 128], [129, 133], [134, 136], [137, 144], [145, 147], [148, 161], [162, 166], [166, 167]]}
{"doc_key": "ai-dev-33", "ner": [[4, 5, "algorithm"], [15, 16, "field"], [18, 19, "researcher"], [21, 22, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[4, 5, 15, 16, "opposite", "", false, false], [18, 19, 15, 16, "related-to", "works_with", false, false], [21, 22, 15, 16, "related-to", "works_with", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["La", "recherche", "sur", "les", "r\u00e9seaux", "neuronaux", "a", "stagn\u00e9", "apr\u00e8s", "la", "publication", "des", "recherches", "sur", "l'", "apprentissage", "automatique", "par", "Marvin", "Minsky", "et", "Seymour", "Papert", "(", "1969", ")", "."], "sentence-detokenized": "La recherche sur les r\u00e9seaux neuronaux a stagn\u00e9 apr\u00e8s la publication des recherches sur l'apprentissage automatique par Marvin Minsky et Seymour Papert (1969).", "token2charspan": [[0, 2], [3, 12], [13, 16], [17, 20], [21, 28], [29, 38], [39, 40], [41, 47], [48, 53], [54, 56], [57, 68], [69, 72], [73, 83], [84, 87], [88, 90], [90, 103], [104, 115], [116, 119], [120, 126], [127, 133], [134, 136], [137, 144], [145, 151], [152, 153], [153, 157], [157, 158], [158, 159]]}
{"doc_key": "ai-dev-34", "ner": [[18, 19, "organisation"], [21, 21, "organisation"], [25, 25, "country"], [27, 27, "country"], [28, 31, "organisation"], [35, 35, "country"], [36, 37, "organisation"], [41, 41, "country"], [42, 42, "organisation"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "relations": [[28, 31, 25, 25, "general-affiliation", "", false, false], [28, 31, 27, 27, "general-affiliation", "", false, false], [36, 37, 35, 35, "general-affiliation", "", false, false], [42, 42, 41, 41, "general-affiliation", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Seules", "quelques", "entreprises", "non", "japonaises", "ont", "finalement", "r\u00e9ussi", "\u00e0", "survivre", "sur", "ce", "march\u00e9", ",", "les", "principales", "\u00e9tant", ":", "Adept", "Technology", ",", "St\u00e4ubli", ",", "la", "soci\u00e9t\u00e9", "su\u00e9doise", "et", "suisse", "ABB", "Asea", "Brown", "Boveri", ",", "la", "soci\u00e9t\u00e9", "allemande", "KUKA", "Robotics", "et", "la", "soci\u00e9t\u00e9", "italienne", "Comau", "."], "sentence-detokenized": "Seules quelques entreprises non japonaises ont finalement r\u00e9ussi \u00e0 survivre sur ce march\u00e9, les principales \u00e9tant : Adept Technology, St\u00e4ubli, la soci\u00e9t\u00e9 su\u00e9doise et suisse ABB Asea Brown Boveri, la soci\u00e9t\u00e9 allemande KUKA Robotics et la soci\u00e9t\u00e9 italienne Comau.", "token2charspan": [[0, 6], [7, 15], [16, 27], [28, 31], [32, 42], [43, 46], [47, 57], [58, 64], [65, 66], [67, 75], [76, 79], [80, 82], [83, 89], [89, 90], [91, 94], [95, 106], [107, 112], [113, 114], [115, 120], [121, 131], [131, 132], [133, 140], [140, 141], [142, 144], [145, 152], [153, 161], [162, 164], [165, 171], [172, 175], [176, 180], [181, 186], [187, 193], [193, 194], [195, 197], [198, 205], [206, 215], [216, 220], [221, 229], [230, 232], [233, 235], [236, 243], [244, 253], [254, 259], [259, 260]]}
{"doc_key": "ai-dev-35", "ner": [[12, 13, "conference"], [21, 21, "conference"]], "ner_mapping_to_source": [0, 1], "relations": [[21, 21, 12, 13, "named", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Les", "activit\u00e9s", "de", "recherche", "comprennent", "une", "conf\u00e9rence", "annuelle", "de", "recherche", ",", "le", "symposium", "RuleML", ",", "\u00e9galement", "connu", "sous", "le", "nom", "de", "RuleML", "."], "sentence-detokenized": "Les activit\u00e9s de recherche comprennent une conf\u00e9rence annuelle de recherche, le symposium RuleML, \u00e9galement connu sous le nom de RuleML.", "token2charspan": [[0, 3], [4, 13], [14, 16], [17, 26], [27, 38], [39, 42], [43, 53], [54, 62], [63, 65], [66, 75], [75, 76], [77, 79], [80, 89], [90, 96], [96, 97], [98, 107], [108, 113], [114, 118], [119, 121], [122, 125], [126, 128], [129, 135], [135, 136]]}
{"doc_key": "ai-dev-36", "ner": [[10, 10, "field"], [13, 13, "field"], [17, 19, "field"], [22, 23, "field"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [], "relations_mapping_to_source": [], "sentence": ["Les", "concepts", "sont", "utilis\u00e9s", "comme", "outils", "ou", "mod\u00e8les", "formels", "en", "math\u00e9matiques", ",", "en", "informatique", ",", "dans", "les", "bases", "de", "donn\u00e9es", "et", "en", "intelligence", "artificielle", ",", "o\u00f9", "ils", "sont", "parfois", "appel\u00e9s", "classes", ",", "sch\u00e9mas", "ou", "cat\u00e9gories", "."], "sentence-detokenized": "Les concepts sont utilis\u00e9s comme outils ou mod\u00e8les formels en math\u00e9matiques, en informatique, dans les bases de donn\u00e9es et en intelligence artificielle, o\u00f9 ils sont parfois appel\u00e9s classes, sch\u00e9mas ou cat\u00e9gories.", "token2charspan": [[0, 3], [4, 12], [13, 17], [18, 26], [27, 32], [33, 39], [40, 42], [43, 50], [51, 58], [59, 61], [62, 75], [75, 76], [77, 79], [80, 92], [92, 93], [94, 98], [99, 102], [103, 108], [109, 111], [112, 119], [120, 122], [123, 125], [126, 138], [139, 151], [151, 152], [153, 155], [156, 159], [160, 164], [165, 172], [173, 180], [181, 188], [188, 189], [190, 197], [198, 200], [201, 211], [211, 212]]}
{"doc_key": "ai-dev-37", "ner": [[6, 8, "organisation"], [11, 14, "organisation"], [17, 17, "organisation"], [20, 22, "organisation"], [25, 27, "organisation"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "a", "\u00e9t\u00e9", "r\u00e9compens\u00e9", "par", "l'", "American", "Psychological", "Association", ",", "la", "National", "Academy", "of", "Sciences", ",", "la", "Royal", ",", "la", "Cognitive", "Neuroscience", "Society", "et", "l'", "American", "Humanist", "Association", "."], "sentence-detokenized": "Il a \u00e9t\u00e9 r\u00e9compens\u00e9 par l'American Psychological Association, la National Academy of Sciences, la Royal, la Cognitive Neuroscience Society et l'American Humanist Association.", "token2charspan": [[0, 2], [3, 4], [5, 8], [9, 19], [20, 23], [24, 26], [26, 34], [35, 48], [49, 60], [60, 61], [62, 64], [65, 73], [74, 81], [82, 84], [85, 93], [93, 94], [95, 97], [98, 103], [103, 104], [105, 107], [108, 117], [118, 130], [131, 138], [139, 141], [142, 144], [144, 152], [153, 161], [162, 173], [173, 174]]}
{"doc_key": "ai-dev-38", "ner": [[1, 2, "person"], [4, 5, "person"], [7, 8, "person"], [18, 20, "person"], [21, 26, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[21, 26, 18, 20, "artifact", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Avec", "Harrison", "Ford", ",", "Rutger", "Hauer", "et", "Sean", "Young", ",", "le", "film", "est", "librement", "inspir\u00e9", "du", "roman", "de", "Philip", "K.", "Dick", "Do", "Androids", "Dream", "of", "Electric", "Sheep", "(", "1968", ")", "."], "sentence-detokenized": "Avec Harrison Ford, Rutger Hauer et Sean Young, le film est librement inspir\u00e9 du roman de Philip K. Dick Do Androids Dream of Electric Sheep (1968).", "token2charspan": [[0, 4], [5, 13], [14, 18], [18, 19], [20, 26], [27, 32], [33, 35], [36, 40], [41, 46], [46, 47], [48, 50], [51, 55], [56, 59], [60, 69], [70, 77], [78, 80], [81, 86], [87, 89], [90, 96], [97, 99], [100, 104], [105, 107], [108, 116], [117, 122], [123, 125], [126, 134], [135, 140], [141, 142], [142, 146], [146, 147], [147, 148]]}
{"doc_key": "ai-dev-39", "ner": [[0, 3, "task"], [8, 13, "algorithm"], [20, 22, "field"], [25, 27, "task"], [30, 31, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[0, 3, 8, 13, "usage", "", false, false], [0, 3, 20, 22, "part-of", "task_part_of_field", false, false], [0, 3, 25, 27, "part-of", "task_part_of_field", false, false], [0, 3, 30, 31, "part-of", "task_part_of_field", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["La", "segmentation", "des", "images", "\u00e0", "l'", "aide", "d'", "algorithmes", "de", "regroupement", "de", "type", "k-means", "est", "utilis\u00e9e", "depuis", "longtemps", "pour", "la", "reconnaissance", "des", "formes", ",", "la", "d\u00e9tection", "des", "objets", "et", "l'", "imagerie", "m\u00e9dicale", "."], "sentence-detokenized": "La segmentation des images \u00e0 l'aide d'algorithmes de regroupement de type k-means est utilis\u00e9e depuis longtemps pour la reconnaissance des formes, la d\u00e9tection des objets et l'imagerie m\u00e9dicale.", "token2charspan": [[0, 2], [3, 15], [16, 19], [20, 26], [27, 28], [29, 31], [31, 35], [36, 38], [38, 49], [50, 52], [53, 65], [66, 68], [69, 73], [74, 81], [82, 85], [86, 94], [95, 101], [102, 111], [112, 116], [117, 119], [120, 134], [135, 138], [139, 145], [145, 146], [147, 149], [150, 159], [160, 163], [164, 170], [171, 173], [174, 176], [176, 184], [185, 193], [193, 194]]}
{"doc_key": "ai-dev-40", "ner": [[18, 18, "algorithm"], [23, 24, "algorithm"], [27, 27, "programlang"]], "ner_mapping_to_source": [0, 1, 2], "relations": [], "relations_mapping_to_source": [], "sentence": ["L'", "\u00e9chantillonnage", "g\u00e9n\u00e9ral", "\u00e0", "partir", "de", "la", "normale", "tronqu\u00e9e", "peut", "\u00eatre", "r\u00e9alis\u00e9", "en", "utilisant", "des", "approximations", "de", "la", "CDF", "normale", "et", "de", "la", "fonction", "probit", ",", "et", "R", "poss\u00e8de", "une", "fonction", "codertnorm", "(", ")", "/", "code", "pour", "g\u00e9n\u00e9rer", "des", "\u00e9chantillons", "normaux", "tronqu\u00e9s", "."], "sentence-detokenized": "L'\u00e9chantillonnage g\u00e9n\u00e9ral \u00e0 partir de la normale tronqu\u00e9e peut \u00eatre r\u00e9alis\u00e9 en utilisant des approximations de la CDF normale et de la fonction probit, et R poss\u00e8de une fonction codertnorm () / code pour g\u00e9n\u00e9rer des \u00e9chantillons normaux tronqu\u00e9s.", "token2charspan": [[0, 2], [2, 17], [18, 25], [26, 27], [28, 34], [35, 37], [38, 40], [41, 48], [49, 57], [58, 62], [63, 67], [68, 75], [76, 78], [79, 88], [89, 92], [93, 107], [108, 110], [111, 113], [114, 117], [118, 125], [126, 128], [129, 131], [132, 134], [135, 143], [144, 150], [150, 151], [152, 154], [155, 156], [157, 164], [165, 168], [169, 177], [178, 188], [189, 190], [190, 191], [192, 193], [194, 198], [199, 203], [204, 211], [212, 215], [216, 228], [229, 236], [237, 245], [245, 246]]}
{"doc_key": "ai-dev-41", "ner": [[9, 11, "university"], [14, 14, "university"], [18, 21, "university"], [25, 27, "university"], [31, 33, "university"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "a", "\u00e9galement", "re\u00e7u", "des", "doctorats", "honorifiques", "de", "l'", "universit\u00e9", "de", "Newcastle", ",", "du", "Surrey", ",", "de", "l'", "universit\u00e9", "de", "Tel", "Aviv", ",", "de", "l'", "universit\u00e9", "Simon", "Fraser", "et", "de", "l'", "universit\u00e9", "de", "Troms\u00f8", "."], "sentence-detokenized": "Il a \u00e9galement re\u00e7u des doctorats honorifiques de l'universit\u00e9 de Newcastle, du Surrey, de l'universit\u00e9 de Tel Aviv, de l'universit\u00e9 Simon Fraser et de l'universit\u00e9 de Troms\u00f8.", "token2charspan": [[0, 2], [3, 4], [5, 14], [15, 19], [20, 23], [24, 33], [34, 46], [47, 49], [50, 52], [52, 62], [63, 65], [66, 75], [75, 76], [77, 79], [80, 86], [86, 87], [88, 90], [91, 93], [93, 103], [104, 106], [107, 110], [111, 115], [115, 116], [117, 119], [120, 122], [122, 132], [133, 138], [139, 145], [146, 148], [149, 151], [152, 154], [154, 164], [165, 167], [168, 174], [174, 175]]}
{"doc_key": "ai-dev-42", "ner": [[4, 4, "programlang"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Une", "mise", "en", "\u0153uvre", "Java", "utilisant", "des", "index", "de", "tableaux", "bas\u00e9s", "sur", "z\u00e9ro", "ainsi", "qu'", "une", "m\u00e9thode", "pratique", "pour", "imprimer", "l'", "ordre", "des", "op\u00e9rations", "r\u00e9solues", ":"], "sentence-detokenized": "Une mise en \u0153uvre Java utilisant des index de tableaux bas\u00e9s sur z\u00e9ro ainsi qu'une m\u00e9thode pratique pour imprimer l'ordre des op\u00e9rations r\u00e9solues :", "token2charspan": [[0, 3], [4, 8], [9, 11], [12, 17], [18, 22], [23, 32], [33, 36], [37, 42], [43, 45], [46, 54], [55, 60], [61, 64], [65, 69], [70, 75], [76, 79], [79, 82], [83, 90], [91, 99], [100, 104], [105, 113], [114, 116], [116, 121], [122, 125], [126, 136], [137, 145], [146, 147]]}
{"doc_key": "ai-dev-43", "ner": [[9, 10, "metrics"], [13, 13, "metrics"], [25, 27, "algorithm"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[13, 13, 9, 10, "named", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Ces", "r\u00e9seaux", "sont", "g\u00e9n\u00e9ralement", "entra\u00een\u00e9s", "sous", "un", "r\u00e9gime", "d'", "entropie", "crois\u00e9e", "(", "ou", "cross-entropy", ")", ",", "ce", "qui", "donne", "une", "variante", "non", "lin\u00e9aire", "de", "la", "r\u00e9gression", "logistique", "multinomiale", "."], "sentence-detokenized": "Ces r\u00e9seaux sont g\u00e9n\u00e9ralement entra\u00een\u00e9s sous un r\u00e9gime d'entropie crois\u00e9e (ou cross-entropy), ce qui donne une variante non lin\u00e9aire de la r\u00e9gression logistique multinomiale.", "token2charspan": [[0, 3], [4, 11], [12, 16], [17, 29], [30, 39], [40, 44], [45, 47], [48, 54], [55, 57], [57, 65], [66, 73], [74, 75], [75, 77], [78, 91], [91, 92], [92, 93], [94, 96], [97, 100], [101, 106], [107, 110], [111, 119], [120, 123], [124, 132], [133, 135], [136, 138], [139, 149], [150, 160], [161, 173], [173, 174]]}
{"doc_key": "ai-dev-44", "ner": [[1, 1, "conference"], [5, 5, "misc"], [4, 14, "conference"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[5, 5, 1, 1, "part-of", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["L'", "ACL", "a", "un", "chapitre", "europ\u00e9en", "(", "chapitre", "europ\u00e9en", "de", "l'", "Association", "for", "Computational", "Linguistics", ")", "."], "sentence-detokenized": "L'ACL a un chapitre europ\u00e9en (chapitre europ\u00e9en de l'Association for Computational Linguistics).", "token2charspan": [[0, 2], [2, 5], [6, 7], [8, 10], [11, 19], [20, 28], [29, 30], [30, 38], [39, 47], [48, 50], [51, 53], [53, 64], [65, 68], [69, 82], [83, 94], [94, 95], [95, 96]]}
{"doc_key": "ai-dev-45", "ner": [[3, 4, "researcher"], [6, 8, "researcher"], [25, 25, "misc"], [28, 29, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[3, 4, 25, 25, "role", "", false, false], [6, 8, 25, 25, "role", "", false, false], [25, 25, 28, 29, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Deux", "professeurs", ",", "Hal", "Abelson", "et", "Gerald", "Jay", "Sussman", ",", "ont", "choisi", "de", "rester", "neutres", "-", "leur", "groupe", "a", "\u00e9t\u00e9", "d\u00e9sign\u00e9", "sous", "les", "noms", "de", "Suisse", "et", "de", "Projet", "MAC", "pendant", "les", "30", "ann\u00e9es", "suivantes", "."], "sentence-detokenized": "Deux professeurs, Hal Abelson et Gerald Jay Sussman, ont choisi de rester neutres - leur groupe a \u00e9t\u00e9 d\u00e9sign\u00e9 sous les noms de Suisse et de Projet MAC pendant les 30 ann\u00e9es suivantes.", "token2charspan": [[0, 4], [5, 16], [16, 17], [18, 21], [22, 29], [30, 32], [33, 39], [40, 43], [44, 51], [51, 52], [53, 56], [57, 63], [64, 66], [67, 73], [74, 81], [82, 83], [84, 88], [89, 95], [96, 97], [98, 101], [102, 109], [110, 114], [115, 118], [119, 123], [124, 126], [127, 133], [134, 136], [137, 139], [140, 146], [147, 150], [151, 158], [159, 162], [163, 165], [166, 172], [173, 182], [182, 183]]}
{"doc_key": "ai-dev-46", "ner": [[2, 2, "misc"], [4, 5, "researcher"], [8, 10, "university"], [19, 19, "organisation"], [21, 24, "organisation"], [31, 32, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[2, 2, 4, 5, "temporal", "", false, false], [4, 5, 19, 19, "physical", "", false, false], [4, 5, 19, 19, "role", "", false, false], [4, 5, 21, 24, "role", "", false, false], [21, 24, 8, 10, "part-of", "", false, false], [31, 32, 21, 24, "role", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["Apr\u00e8s", "son", "doctorat", ",", "Ghahramani", "a", "rejoint", "l'", "Universit\u00e9", "de", "Toronto", "en", "1995", "en", "tant", "que", "boursier", "postdoctoral", "du", "CIRT", "au", "laboratoire", "d'", "intelligence", "artificielle", ",", "o\u00f9", "il", "a", "travaill\u00e9", "avec", "Geoffrey", "Hinton", "."], "sentence-detokenized": "Apr\u00e8s son doctorat, Ghahramani a rejoint l'Universit\u00e9 de Toronto en 1995 en tant que boursier postdoctoral du CIRT au laboratoire d'intelligence artificielle, o\u00f9 il a travaill\u00e9 avec Geoffrey Hinton.", "token2charspan": [[0, 5], [6, 9], [10, 18], [18, 19], [20, 30], [31, 32], [33, 40], [41, 43], [43, 53], [54, 56], [57, 64], [65, 67], [68, 72], [73, 75], [76, 80], [81, 84], [85, 93], [94, 106], [107, 109], [110, 114], [115, 117], [118, 129], [130, 132], [132, 144], [145, 157], [157, 158], [159, 161], [162, 164], [165, 166], [167, 176], [177, 181], [182, 190], [191, 197], [197, 198]]}
{"doc_key": "ai-dev-47", "ner": [[31, 33, "metrics"], [35, 35, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[35, 35, 31, 33, "named", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Des", "travaux", "ult\u00e9rieurs", "se", "sont", "attach\u00e9s", "\u00e0", "r\u00e9soudre", "ces", "probl\u00e8mes", ",", "mais", "ce", "n'", "est", "qu'", "avec", "l'", "av\u00e8nement", "de", "l'", "ordinateur", "moderne", "et", "la", "popularisation", "des", "techniques", "de", "param\u00e9trage", "par", "Maximum", "de", "vraisemblance", "(", "MLE", ")", "que", "la", "recherche", "a", "r\u00e9ellement", "pris", "son", "essor", "."], "sentence-detokenized": "Des travaux ult\u00e9rieurs se sont attach\u00e9s \u00e0 r\u00e9soudre ces probl\u00e8mes, mais ce n'est qu'avec l'av\u00e8nement de l'ordinateur moderne et la popularisation des techniques de param\u00e9trage par Maximum de vraisemblance (MLE) que la recherche a r\u00e9ellement pris son essor.", "token2charspan": [[0, 3], [4, 11], [12, 22], [23, 25], [26, 30], [31, 39], [40, 41], [42, 50], [51, 54], [55, 64], [64, 65], [66, 70], [71, 73], [74, 76], [76, 79], [80, 83], [83, 87], [88, 90], [90, 99], [100, 102], [103, 105], [105, 115], [116, 123], [124, 126], [127, 129], [130, 144], [145, 148], [149, 159], [160, 162], [163, 174], [175, 178], [179, 186], [187, 189], [190, 203], [204, 205], [205, 208], [208, 209], [210, 213], [214, 216], [217, 226], [227, 228], [229, 239], [240, 244], [245, 248], [249, 254], [254, 255]]}
{"doc_key": "ai-dev-48", "ner": [[6, 7, "person"], [10, 11, "person"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["La", "s\u00e9rie", "a", "\u00e9t\u00e9", "produite", "par", "David", "Fincher", ",", "avec", "Kevin", "Spacey", "en", "vedette", "."], "sentence-detokenized": "La s\u00e9rie a \u00e9t\u00e9 produite par David Fincher, avec Kevin Spacey en vedette.", "token2charspan": [[0, 2], [3, 8], [9, 10], [11, 14], [15, 23], [24, 27], [28, 33], [34, 41], [41, 42], [43, 47], [48, 53], [54, 60], [61, 63], [64, 71], [71, 72]]}
{"doc_key": "ai-dev-49", "ner": [[22, 22, "metrics"], [32, 35, "algorithm"], [42, 43, "algorithm"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[32, 35, 42, 43, "compare", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["En", "raison", "des", "limites", "de", "la", "puissance", "de", "calcul", ",", "les", "m\u00e9thodes", "in", "silico", "actuelles", "doivent", "g\u00e9n\u00e9ralement", "troquer", "la", "vitesse", "contre", "la", "pr\u00e9cision", ";", "par", "exemple", ",", "utiliser", "des", "m\u00e9thodes", "rapides", "d'", "arrimage", "des", "prot\u00e9ines", "au", "lieu", "de", "calculs", "co\u00fbteux", "de", "l'", "\u00e9nergie", "libre", "."], "sentence-detokenized": "En raison des limites de la puissance de calcul, les m\u00e9thodes in silico actuelles doivent g\u00e9n\u00e9ralement troquer la vitesse contre la pr\u00e9cision ; par exemple, utiliser des m\u00e9thodes rapides d'arrimage des prot\u00e9ines au lieu de calculs co\u00fbteux de l'\u00e9nergie libre.", "token2charspan": [[0, 2], [3, 9], [10, 13], [14, 21], [22, 24], [25, 27], [28, 37], [38, 40], [41, 47], [47, 48], [49, 52], [53, 61], [62, 64], [65, 71], [72, 81], [82, 89], [90, 102], [103, 110], [111, 113], [114, 121], [122, 128], [129, 131], [132, 141], [142, 143], [144, 147], [148, 155], [155, 156], [157, 165], [166, 169], [170, 178], [179, 186], [187, 189], [189, 197], [198, 201], [202, 211], [212, 214], [215, 219], [220, 222], [223, 230], [231, 238], [239, 241], [242, 244], [244, 251], [252, 257], [257, 258]]}
{"doc_key": "ai-dev-50", "ner": [[7, 7, "country"], [9, 10, "country"], [13, 13, "country"], [16, 16, "country"], [19, 19, "country"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [], "relations_mapping_to_source": [], "sentence": ["Elle", "comptait", "plus", "de", "30", "sites", "aux", "\u00c9tats-Unis", ",", "au", "Canada", ",", "au", "Mexique", ",", "au", "Br\u00e9sil", "et", "en", "Argentine", "."], "sentence-detokenized": "Elle comptait plus de 30 sites aux \u00c9tats-Unis, au Canada, au Mexique, au Br\u00e9sil et en Argentine.", "token2charspan": [[0, 4], [5, 13], [14, 18], [19, 21], [22, 24], [25, 30], [31, 34], [35, 45], [45, 46], [47, 49], [50, 56], [56, 57], [58, 60], [61, 68], [68, 69], [70, 72], [73, 79], [80, 82], [83, 85], [86, 95], [95, 96]]}
{"doc_key": "ai-dev-51", "ner": [[7, 9, "field"], [13, 16, "product"], [18, 20, "algorithm"], [29, 31, "task"], [34, 36, "task"], [41, 41, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[13, 16, 7, 9, "part-of", "", false, false], [13, 16, 18, 20, "usage", "", false, false], [29, 31, 7, 9, "part-of", "task_part_of_field", false, false], [29, 31, 41, 41, "related-to", "performs", false, false], [34, 36, 7, 9, "part-of", "task_part_of_field", false, false], [34, 36, 41, 41, "related-to", "performs", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["Un", "exemple", "de", "pipeline", "de", "calcul", "de", "vision", "par", "ordinateur", "typique", "pour", "un", "syst\u00e8me", "de", "reconnaissance", "faciale", "utilisant", "k", "-", "NN", ",", "y", "compris", "les", "\u00e9tapes", "de", "pr\u00e9traitement", "d'", "extraction", "de", "caract\u00e9ristiques", "et", "de", "r\u00e9duction", "de", "dimension", "(", "g\u00e9n\u00e9ralement", "impl\u00e9ment\u00e9", "avec", "OpenCV", ")", ":"], "sentence-detokenized": "Un exemple de pipeline de calcul de vision par ordinateur typique pour un syst\u00e8me de reconnaissance faciale utilisant k -NN, y compris les \u00e9tapes de pr\u00e9traitement d'extraction de caract\u00e9ristiques et de r\u00e9duction de dimension (g\u00e9n\u00e9ralement impl\u00e9ment\u00e9 avec OpenCV) :", "token2charspan": [[0, 2], [3, 10], [11, 13], [14, 22], [23, 25], [26, 32], [33, 35], [36, 42], [43, 46], [47, 57], [58, 65], [66, 70], [71, 73], [74, 81], [82, 84], [85, 99], [100, 107], [108, 117], [118, 119], [120, 121], [121, 123], [123, 124], [125, 126], [127, 134], [135, 138], [139, 145], [146, 148], [149, 162], [163, 165], [165, 175], [176, 178], [179, 195], [196, 198], [199, 201], [202, 211], [212, 214], [215, 224], [225, 226], [226, 238], [239, 249], [250, 254], [255, 261], [261, 262], [263, 264]]}
{"doc_key": "ai-dev-52", "ner": [[13, 17, "algorithm"], [20, 20, "misc"], [23, 24, "misc"], [28, 29, "misc"], [35, 35, "programlang"], [37, 37, "product"], [43, 44, "algorithm"], [48, 49, "misc"], [52, 52, "misc"], [54, 54, "misc"], [56, 56, "misc"], [66, 66, "misc"], [69, 75, "misc"], [72, 74, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "dispose", "d'", "un", "riche", "ensemble", "de", "fonctionnalit\u00e9s", ",", "de", "biblioth\u00e8ques", "pour", "la", "programmation", "en", "logique", "des", "contraintes", ",", "du", "multithreading", ",", "des", "tests", "unitaires", ",", "d'", "une", "interface", "graphique", ",", "de", "l'", "interfa\u00e7age", "avec", "Java", ",", "ODBC", "et", "autres", ",", "de", "la", "programmation", "lettr\u00e9e", ",", "d'", "un", "serveur", "web", ",", "de", "SGML", ",", "RDF", ",", "RDFS", ",", "d'", "outils", "de", "d\u00e9veloppement", "(", "y", "compris", "un", "IDE", "avec", "un", "d\u00e9bogueur", "et", "un", "profileur", "d'", "interface", "graphique", ")", "et", "d'", "une", "documentation", "\u00e9tendue", "."], "sentence-detokenized": "Il dispose d'un riche ensemble de fonctionnalit\u00e9s, de biblioth\u00e8ques pour la programmation en logique des contraintes, du multithreading, des tests unitaires, d'une interface graphique, de l'interfa\u00e7age avec Java, ODBC et autres, de la programmation lettr\u00e9e, d'un serveur web, de SGML, RDF, RDFS, d'outils de d\u00e9veloppement (y compris un IDE avec un d\u00e9bogueur et un profileur d'interface graphique) et d'une documentation \u00e9tendue.", "token2charspan": [[0, 2], [3, 10], [11, 13], [13, 15], [16, 21], [22, 30], [31, 33], [34, 49], [49, 50], [51, 53], [54, 67], [68, 72], [73, 75], [76, 89], [90, 92], [93, 100], [101, 104], [105, 116], [116, 117], [118, 120], [121, 135], [135, 136], [137, 140], [141, 146], [147, 156], [156, 157], [158, 160], [160, 163], [164, 173], [174, 183], [183, 184], [185, 187], [188, 190], [190, 201], [202, 206], [207, 211], [211, 212], [213, 217], [218, 220], [221, 227], [227, 228], [229, 231], [232, 234], [235, 248], [249, 256], [256, 257], [258, 260], [260, 262], [263, 270], [271, 274], [274, 275], [276, 278], [279, 283], [283, 284], [285, 288], [288, 289], [290, 294], [294, 295], [296, 298], [298, 304], [305, 307], [308, 321], [322, 323], [323, 324], [325, 332], [333, 335], [336, 339], [340, 344], [345, 347], [348, 357], [358, 360], [361, 363], [364, 373], [374, 376], [376, 385], [386, 395], [395, 396], [397, 399], [400, 402], [402, 405], [406, 419], [420, 427], [427, 428]]}
{"doc_key": "ai-dev-53", "ner": [[5, 7, "field"], [10, 12, "field"], [17, 22, "misc"], [25, 30, "misc"], [32, 34, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[17, 22, 5, 7, "part-of", "", true, false], [17, 22, 10, 12, "part-of", "", false, false], [17, 22, 32, 34, "type-of", "", false, false], [25, 30, 5, 7, "part-of", "", false, false], [25, 30, 10, 12, "part-of", "", false, false], [25, 30, 32, 34, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["Dans", "le", "domaine", "de", "la", "vision", "par", "ordinateur", "et", "du", "traitement", "des", "images", ",", "la", "notion", "de", "repr\u00e9sentation", "de", "l'", "espace", "d'", "\u00e9chelle", "et", "les", "op\u00e9rateurs", "de", "la", "d\u00e9riv\u00e9e", "gaussienne", "constituent", "une", "repr\u00e9sentation", "multi-\u00e9chelle", "canonique", "."], "sentence-detokenized": "Dans le domaine de la vision par ordinateur et du traitement des images, la notion de repr\u00e9sentation de l'espace d'\u00e9chelle et les op\u00e9rateurs de la d\u00e9riv\u00e9e gaussienne constituent une repr\u00e9sentation multi-\u00e9chelle canonique.", "token2charspan": [[0, 4], [5, 7], [8, 15], [16, 18], [19, 21], [22, 28], [29, 32], [33, 43], [44, 46], [47, 49], [50, 60], [61, 64], [65, 71], [71, 72], [73, 75], [76, 82], [83, 85], [86, 100], [101, 103], [104, 106], [106, 112], [113, 115], [115, 122], [123, 125], [126, 129], [130, 140], [141, 143], [144, 146], [147, 154], [155, 165], [166, 177], [178, 181], [182, 196], [197, 210], [211, 220], [220, 221]]}
{"doc_key": "ai-dev-54", "ner": [[7, 11, "organisation"], [22, 32, "conference"]], "ner_mapping_to_source": [0, 1], "relations": [[7, 11, 22, 32, "role", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Il", "est", "\u00e9galement", "le", "pr\u00e9sident", "de", "la", "Neural", "Information", "Processing", "Systems", "Foundation", ",", "une", "organisation", "\u00e0", "but", "non", "lucratif", "qui", "supervise", "la", "conf\u00e9rence", "annuelle", "sur", "les", "syst\u00e8mes", "de", "traitement", "de", "l'", "information", "neuronale", "."], "sentence-detokenized": "Il est \u00e9galement le pr\u00e9sident de la Neural Information Processing Systems Foundation, une organisation \u00e0 but non lucratif qui supervise la conf\u00e9rence annuelle sur les syst\u00e8mes de traitement de l'information neuronale.", "token2charspan": [[0, 2], [3, 6], [7, 16], [17, 19], [20, 29], [30, 32], [33, 35], [36, 42], [43, 54], [55, 65], [66, 73], [74, 84], [84, 85], [86, 89], [90, 102], [103, 104], [105, 108], [109, 112], [113, 121], [122, 125], [126, 135], [136, 138], [139, 149], [150, 158], [159, 162], [163, 166], [167, 175], [176, 178], [179, 189], [190, 192], [193, 195], [195, 206], [207, 216], [216, 217]]}
{"doc_key": "ai-dev-55", "ner": [[4, 6, "task"], [9, 10, "metrics"], [15, 17, "misc"], [21, 21, "task"], [24, 25, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[4, 6, 9, 10, "usage", "", false, false], [9, 10, 15, 17, "type-of", "", false, false], [21, 21, 24, 25, "usage", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Pour", "les", "probl\u00e8mes", "d'", "analyse", "de", "r\u00e9gression", ",", "l'", "erreur", "quadratique", "peut", "\u00eatre", "utilis\u00e9e", "comme", "fonction", "de", "perte", ",", "pour", "la", "classification", ",", "l'", "entropie", "crois\u00e9e", "peut", "\u00eatre", "utilis\u00e9e", "."], "sentence-detokenized": "Pour les probl\u00e8mes d'analyse de r\u00e9gression, l'erreur quadratique peut \u00eatre utilis\u00e9e comme fonction de perte, pour la classification, l'entropie crois\u00e9e peut \u00eatre utilis\u00e9e.", "token2charspan": [[0, 4], [5, 8], [9, 18], [19, 21], [21, 28], [29, 31], [32, 42], [42, 43], [44, 46], [46, 52], [53, 64], [65, 69], [70, 74], [75, 83], [84, 89], [90, 98], [99, 101], [102, 107], [107, 108], [109, 113], [114, 116], [117, 131], [131, 132], [133, 135], [135, 143], [144, 151], [152, 156], [157, 161], [162, 170], [170, 171]]}
{"doc_key": "ai-dev-56", "ner": [[0, 1, "researcher"], [24, 37, "conference"], [29, 38, "conference"], [54, 54, "university"], [50, 52, "field"], [63, 67, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[0, 1, 24, 37, "role", "", false, false], [0, 1, 54, 54, "physical", "", false, false], [0, 1, 54, 54, "role", "", false, false], [0, 1, 63, 67, "role", "", false, false], [24, 37, 29, 38, "named", "same", false, false], [54, 54, 50, 52, "related-to", "subject_of_study_at", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["M.", "Lafferty", "a", "occup\u00e9", "de", "nombreux", "postes", "prestigieux", ",", "notamment", ":", "1", ")", "copr\u00e9sident", "du", "programme", "et", "copr\u00e9sident", "g\u00e9n\u00e9ral", "des", "conf\u00e9rences", "de", "la", "fondation", "Neural", "Information", "Processing", "Systems", "(", "conf\u00e9rence", "sur", "les", "syst\u00e8mes", "de", "traitement", "de", "l'", "information", "neuronale", ")", ";", "2", ")", "codirecteur", "du", "nouveau", "programme", "de", "doctorat", "en", "apprentissage", "automatique", "de", "la", "CMU", ";", "3", ")", "r\u00e9dacteur", "en", "chef", "adjoint", "du", "Journal", "of", "Machine", "Learning", "Research", "."], "sentence-detokenized": "M. Lafferty a occup\u00e9 de nombreux postes prestigieux, notamment : 1) copr\u00e9sident du programme et copr\u00e9sident g\u00e9n\u00e9ral des conf\u00e9rences de la fondation Neural Information Processing Systems (conf\u00e9rence sur les syst\u00e8mes de traitement de l'information neuronale) ; 2) codirecteur du nouveau programme de doctorat en apprentissage automatique de la CMU ; 3) r\u00e9dacteur en chef adjoint du Journal of Machine Learning Research.", "token2charspan": [[0, 2], [3, 11], [12, 13], [14, 20], [21, 23], [24, 32], [33, 39], [40, 51], [51, 52], [53, 62], [63, 64], [65, 66], [66, 67], [68, 79], [80, 82], [83, 92], [93, 95], [96, 107], [108, 115], [116, 119], [120, 131], [132, 134], [135, 137], [138, 147], [148, 154], [155, 166], [167, 177], [178, 185], [186, 187], [187, 197], [198, 201], [202, 205], [206, 214], [215, 217], [218, 228], [229, 231], [232, 234], [234, 245], [246, 255], [255, 256], [257, 258], [259, 260], [260, 261], [262, 273], [274, 276], [277, 284], [285, 294], [295, 297], [298, 306], [307, 309], [310, 323], [324, 335], [336, 338], [339, 341], [342, 345], [346, 347], [348, 349], [349, 350], [351, 360], [361, 363], [364, 368], [369, 376], [377, 379], [380, 387], [388, 390], [391, 398], [399, 407], [408, 416], [416, 417]]}
{"doc_key": "ai-dev-57", "ner": [[0, 2, "misc"], [6, 6, "algorithm"], [8, 8, "algorithm"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[6, 6, 0, 2, "type-of", "", false, false], [8, 8, 0, 2, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Les", "algorithmes", "convexes", ",", "tels", "que", "AdaBoost", "et", "LogitBoost", ",", "peuvent", "\u00eatre", "mis", "en", "\u00e9chec", "par", "le", "bruit", "al\u00e9atoire", ",", "de", "sorte", "qu'", "ils", "ne", "peuvent", "pas", "apprendre", "les", "combinaisons", "de", "base", "et", "apprenables", "d'", "hypoth\u00e8ses", "faibles", "."], "sentence-detokenized": "Les algorithmes convexes, tels que AdaBoost et LogitBoost, peuvent \u00eatre mis en \u00e9chec par le bruit al\u00e9atoire, de sorte qu'ils ne peuvent pas apprendre les combinaisons de base et apprenables d'hypoth\u00e8ses faibles.", "token2charspan": [[0, 3], [4, 15], [16, 24], [24, 25], [26, 30], [31, 34], [35, 43], [44, 46], [47, 57], [57, 58], [59, 66], [67, 71], [72, 75], [76, 78], [79, 84], [85, 88], [89, 91], [92, 97], [98, 107], [107, 108], [109, 111], [112, 117], [118, 121], [121, 124], [125, 127], [128, 135], [136, 139], [140, 149], [150, 153], [154, 166], [167, 169], [170, 174], [175, 177], [178, 189], [190, 192], [192, 202], [203, 210], [210, 211]]}
{"doc_key": "ai-dev-58", "ner": [[0, 0, "product"], [3, 9, "product"], [14, 17, "algorithm"], [27, 29, "algorithm"], [32, 36, "task"], [39, 43, "task"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[0, 0, 3, 9, "type-of", "", false, false], [0, 0, 14, 17, "usage", "", false, false], [0, 0, 27, 29, "usage", "", false, false], [27, 29, 32, 36, "related-to", "used_for", true, false], [27, 29, 39, 43, "related-to", "used_for", true, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Apertium", "est", "un", "syst\u00e8me", "de", "traduction", "automatique", "\u00e0", "transfert", "superficiel", ",", "qui", "utilise", "des", "transducteurs", "\u00e0", "\u00e9tats", "finis", "pour", "toutes", "ses", "transformations", "lexicales", ",", "et", "des", "mod\u00e8les", "de", "Markov", "cach\u00e9s", "pour", "le", "marquage", "des", "parties", "du", "discours", "ou", "la", "d\u00e9sambigu\u00efsation", "des", "cat\u00e9gories", "de", "mots", "."], "sentence-detokenized": "Apertium est un syst\u00e8me de traduction automatique \u00e0 transfert superficiel, qui utilise des transducteurs \u00e0 \u00e9tats finis pour toutes ses transformations lexicales, et des mod\u00e8les de Markov cach\u00e9s pour le marquage des parties du discours ou la d\u00e9sambigu\u00efsation des cat\u00e9gories de mots.", "token2charspan": [[0, 8], [9, 12], [13, 15], [16, 23], [24, 26], [27, 37], [38, 49], [50, 51], [52, 61], [62, 73], [73, 74], [75, 78], [79, 86], [87, 90], [91, 104], [105, 106], [107, 112], [113, 118], [119, 123], [124, 130], [131, 134], [135, 150], [151, 160], [160, 161], [162, 164], [165, 168], [169, 176], [177, 179], [180, 186], [187, 193], [194, 198], [199, 201], [202, 210], [211, 214], [215, 222], [223, 225], [226, 234], [235, 237], [238, 240], [241, 257], [258, 261], [262, 272], [273, 275], [276, 280], [280, 281]]}
{"doc_key": "ai-dev-59", "ner": [[1, 2, "misc"], [15, 19, "metrics"], [36, 37, "metrics"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[1, 2, 15, 19, "related-to", "", true, false], [15, 19, 36, 37, "related-to", "", true, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Le", "gradient", "naturel", "de", "mathE", "f", "(", "x", ")", "/", "math", ",", "conforme", "\u00e0", "la", "m\u00e9trique", "d'", "information", "de", "Fisher", "(", "une", "mesure", "de", "distance", "informationnelle", "entre", "des", "distributions", "de", "probabilit\u00e9", "et", "la", "courbure", "de", "l'", "entropie", "relative", ")", ",", "se", "lit", "maintenant", "comme", "suit"], "sentence-detokenized": "Le gradient naturel de mathE f (x) / math, conforme \u00e0 la m\u00e9trique d'information de Fisher (une mesure de distance informationnelle entre des distributions de probabilit\u00e9 et la courbure de l'entropie relative), se lit maintenant comme suit", "token2charspan": [[0, 2], [3, 11], [12, 19], [20, 22], [23, 28], [29, 30], [31, 32], [32, 33], [33, 34], [35, 36], [37, 41], [41, 42], [43, 51], [52, 53], [54, 56], [57, 65], [66, 68], [68, 79], [80, 82], [83, 89], [90, 91], [91, 94], [95, 101], [102, 104], [105, 113], [114, 130], [131, 136], [137, 140], [141, 154], [155, 157], [158, 169], [170, 172], [173, 175], [176, 184], [185, 187], [188, 190], [190, 198], [199, 207], [207, 208], [208, 209], [210, 212], [213, 216], [217, 227], [228, 233], [234, 238]]}
{"doc_key": "ai-dev-60", "ner": [[1, 4, "programlang"], [9, 9, "product"], [11, 11, "programlang"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[9, 9, 1, 4, "origin", "", false, false], [11, 11, 1, 4, "origin", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Le", "langage", "de", "programmation", "S", "a", "inspir\u00e9", "les", "syst\u00e8mes", "S'-PLUS", "et", "R."], "sentence-detokenized": "Le langage de programmation S a inspir\u00e9 les syst\u00e8mes S'-PLUS et R.", "token2charspan": [[0, 2], [3, 10], [11, 13], [14, 27], [28, 29], [30, 31], [32, 39], [40, 43], [44, 52], [53, 60], [61, 63], [64, 66]]}
{"doc_key": "ai-dev-61", "ner": [[8, 8, "product"], [14, 14, "product"], [17, 17, "product"], [23, 25, "researcher"], [27, 28, "researcher"], [30, 31, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[8, 8, 14, 14, "named", "same", false, false], [17, 17, 14, 14, "origin", "derived_from", false, false], [17, 17, 23, 25, "origin", "", false, false], [17, 17, 27, 28, "origin", "", false, false], [17, 17, 30, 31, "origin", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["La", "mise", "en", "\u0153uvre", "la", "plus", "influente", "de", "Planner", "a", "\u00e9t\u00e9", "le", "sous-ensemble", "de", "Planner", ",", "appel\u00e9", "Micro-Planner", ",", "mis", "en", "\u0153uvre", "par", "Gerald", "Jay", "Sussman", ",", "Eugene", "Charniak", "et", "Terry", "Winograd", "."], "sentence-detokenized": "La mise en \u0153uvre la plus influente de Planner a \u00e9t\u00e9 le sous-ensemble de Planner, appel\u00e9 Micro-Planner, mis en \u0153uvre par Gerald Jay Sussman, Eugene Charniak et Terry Winograd.", "token2charspan": [[0, 2], [3, 7], [8, 10], [11, 16], [17, 19], [20, 24], [25, 34], [35, 37], [38, 45], [46, 47], [48, 51], [52, 54], [55, 68], [69, 71], [72, 79], [79, 80], [81, 87], [88, 101], [101, 102], [103, 106], [107, 109], [110, 115], [116, 119], [120, 126], [127, 130], [131, 138], [138, 139], [140, 146], [147, 155], [156, 158], [159, 164], [165, 173], [173, 174]]}
{"doc_key": "ai-dev-62", "ner": [[5, 5, "country"], [6, 9, "researcher"], [22, 22, "misc"], [20, 27, "university"], [36, 37, "misc"], [44, 44, "misc"], [51, 53, "misc"]], "ner_mapping_to_source": [1, 2, 3, 4, 5, 6, 7], "relations": [[6, 9, 5, 5, "general-affiliation", "from_country", false, false], [20, 27, 22, 22, "general-affiliation", "nationality", false, false]], "relations_mapping_to_source": [1, 2], "sentence": ["En", "1779", ",", "le", "scientifique", "germano-danois", "Christian", "Gottlieb", "Kratzenstein", "a", "remport\u00e9", "le", "premier", "prix", "d'", "un", "concours", "annonc\u00e9", "par", "l'", "Acad\u00e9mie", "imp\u00e9riale", "russe", "des", "sciences", "et", "des", "arts", "pour", "les", "mod\u00e8les", "qu'", "il", "a", "construits", "du", "conduit", "vocal", "humain", "capable", "de", "produire", "les", "cinq", "voyelles", "longues", "(", "en", "notation", "de", "l'", "Alphabet", "phon\u00e9tique", "international", ":"], "sentence-detokenized": "En 1779, le scientifique germano-danois Christian Gottlieb Kratzenstein a remport\u00e9 le premier prix d'un concours annonc\u00e9 par l'Acad\u00e9mie imp\u00e9riale russe des sciences et des arts pour les mod\u00e8les qu'il a construits du conduit vocal humain capable de produire les cinq voyelles longues (en notation de l'Alphabet phon\u00e9tique international :", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 11], [12, 24], [25, 39], [40, 49], [50, 58], [59, 71], [72, 73], [74, 82], [83, 85], [86, 93], [94, 98], [99, 101], [101, 103], [104, 112], [113, 120], [121, 124], [125, 127], [127, 135], [136, 145], [146, 151], [152, 155], [156, 164], [165, 167], [168, 171], [172, 176], [177, 181], [182, 185], [186, 193], [194, 197], [197, 199], [200, 201], [202, 212], [213, 215], [216, 223], [224, 229], [230, 236], [237, 244], [245, 247], [248, 256], [257, 260], [261, 265], [266, 274], [275, 282], [283, 284], [284, 286], [287, 295], [296, 298], [299, 301], [301, 309], [310, 320], [321, 334], [335, 336]]}
{"doc_key": "ai-dev-63", "ner": [[5, 6, "product"], [10, 11, "misc"], [14, 20, "misc"], [41, 45, "misc"], [79, 83, "task"], [90, 91, "product"], [94, 94, "product"], [101, 105, "task"], [108, 109, "task"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "relations": [[5, 6, 90, 91, "related-to", "supports_program", false, false], [5, 6, 94, 94, "related-to", "supports_program", false, false], [10, 11, 5, 6, "part-of", "", false, false], [14, 20, 5, 6, "part-of", "", false, false], [41, 45, 5, 6, "part-of", "", false, false], [79, 83, 5, 6, "part-of", "", false, false], [101, 105, 5, 6, "part-of", "", false, false], [108, 109, 5, 6, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "sentence": ["Parmi", "les", "nouvelles", "fonctionnalit\u00e9s", "d'", "Office", "XP", ",", "citons", "les", "balises", "intelligentes", ",", "une", "fonction", "de", "recherche", "bas\u00e9e", "sur", "la", "s\u00e9lection", "qui", "reconna\u00eet", "diff\u00e9rents", "types", "de", "texte", "dans", "un", "document", "afin", "que", "les", "utilisateurs", "puissent", "effectuer", "des", "actions", "suppl\u00e9mentaires", ";", "une", "interface", "de", "volet", "de", "t\u00e2ches", "qui", "regroupe", "les", "commandes", "les", "plus", "courantes", "de", "la", "barre", "de", "menus", "sur", "le", "c\u00f4t\u00e9", "droit", "de", "l'", "\u00e9cran", "afin", "de", "faciliter", "l'", "acc\u00e8s", "rapide", "\u00e0", "celles", "-ci", ";", "de", "nouvelles", "capacit\u00e9s", "de", "collaboration", "en", "mati\u00e8re", "de", "documents", ",", "la", "prise", "en", "charge", "de", "MSN", "Groups", "et", "de", "SharePoint", ";", "et", "des", "capacit\u00e9s", "int\u00e9gr\u00e9es", "de", "reconnaissance", "de", "l'", "\u00e9criture", "manuscrite", "et", "de", "reconnaissance", "vocale", "."], "sentence-detokenized": "Parmi les nouvelles fonctionnalit\u00e9s d'Office XP, citons les balises intelligentes, une fonction de recherche bas\u00e9e sur la s\u00e9lection qui reconna\u00eet diff\u00e9rents types de texte dans un document afin que les utilisateurs puissent effectuer des actions suppl\u00e9mentaires ; une interface de volet de t\u00e2ches qui regroupe les commandes les plus courantes de la barre de menus sur le c\u00f4t\u00e9 droit de l'\u00e9cran afin de faciliter l'acc\u00e8s rapide \u00e0 celles-ci ; de nouvelles capacit\u00e9s de collaboration en mati\u00e8re de documents, la prise en charge de MSN Groups et de SharePoint ; et des capacit\u00e9s int\u00e9gr\u00e9es de reconnaissance de l'\u00e9criture manuscrite et de reconnaissance vocale.", "token2charspan": [[0, 5], [6, 9], [10, 19], [20, 35], [36, 38], [38, 44], [45, 47], [47, 48], [49, 55], [56, 59], [60, 67], [68, 81], [81, 82], [83, 86], [87, 95], [96, 98], [99, 108], [109, 114], [115, 118], [119, 121], [122, 131], [132, 135], [136, 145], [146, 156], [157, 162], [163, 165], [166, 171], [172, 176], [177, 179], [180, 188], [189, 193], [194, 197], [198, 201], [202, 214], [215, 223], [224, 233], [234, 237], [238, 245], [246, 261], [262, 263], [264, 267], [268, 277], [278, 280], [281, 286], [287, 289], [290, 296], [297, 300], [301, 309], [310, 313], [314, 323], [324, 327], [328, 332], [333, 342], [343, 345], [346, 348], [349, 354], [355, 357], [358, 363], [364, 367], [368, 370], [371, 375], [376, 381], [382, 384], [385, 387], [387, 392], [393, 397], [398, 400], [401, 410], [411, 413], [413, 418], [419, 425], [426, 427], [428, 434], [434, 437], [438, 439], [440, 442], [443, 452], [453, 462], [463, 465], [466, 479], [480, 482], [483, 490], [491, 493], [494, 503], [503, 504], [505, 507], [508, 513], [514, 516], [517, 523], [524, 526], [527, 530], [531, 537], [538, 540], [541, 543], [544, 554], [555, 556], [557, 559], [560, 563], [564, 573], [574, 583], [584, 586], [587, 601], [602, 604], [605, 607], [607, 615], [616, 626], [627, 629], [630, 632], [633, 647], [648, 654], [654, 655]]}
{"doc_key": "ai-dev-64", "ner": [[12, 13, "algorithm"], [15, 17, "misc"]], "ner_mapping_to_source": [0, 1], "relations": [[12, 13, 15, 17, "type-of", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Dans", "de", "nombreuses", "applications", ",", "les", "unit\u00e9s", "de", "ces", "r\u00e9seaux", "appliquent", "une", "fonction", "sigmo\u00efde", "comme", "fonction", "d'", "activation", "."], "sentence-detokenized": "Dans de nombreuses applications, les unit\u00e9s de ces r\u00e9seaux appliquent une fonction sigmo\u00efde comme fonction d'activation.", "token2charspan": [[0, 4], [5, 7], [8, 18], [19, 31], [31, 32], [33, 36], [37, 43], [44, 46], [47, 50], [51, 58], [59, 69], [70, 73], [74, 82], [83, 91], [92, 97], [98, 106], [107, 109], [109, 119], [119, 120]]}
{"doc_key": "ai-dev-65", "ner": [[3, 4, "researcher"], [12, 18, "organisation"], [31, 37, "organisation"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[3, 4, 12, 18, "role", "", false, false], [3, 4, 31, 37, "role", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["En", "2001", ",", "Mehler", "a", "\u00e9t\u00e9", "\u00e9lu", "membre", "honoraire", "\u00e9tranger", "de", "l'", "Acad\u00e9mie", "am\u00e9ricaine", "des", "arts", "et", "des", "sciences", ",", "et", "en", "2003", ",", "il", "a", "\u00e9t\u00e9", "\u00e9lu", "membre", "de", "l'", "Association", "am\u00e9ricaine", "pour", "l'", "avancement", "des", "sciences", "."], "sentence-detokenized": "En 2001, Mehler a \u00e9t\u00e9 \u00e9lu membre honoraire \u00e9tranger de l'Acad\u00e9mie am\u00e9ricaine des arts et des sciences, et en 2003, il a \u00e9t\u00e9 \u00e9lu membre de l'Association am\u00e9ricaine pour l'avancement des sciences.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 15], [16, 17], [18, 21], [22, 25], [26, 32], [33, 42], [43, 51], [52, 54], [55, 57], [57, 65], [66, 76], [77, 80], [81, 85], [86, 88], [89, 92], [93, 101], [101, 102], [103, 105], [106, 108], [109, 113], [113, 114], [115, 117], [118, 119], [120, 123], [124, 127], [128, 134], [135, 137], [138, 140], [140, 151], [152, 162], [163, 167], [168, 170], [170, 180], [181, 184], [185, 193], [193, 194]]}
{"doc_key": "ai-dev-66", "ner": [[6, 8, "task"], [11, 13, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[6, 8, 11, 13, "cause-effect", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["L'", "extension", "de", "ce", "concept", "aux", "classifications", "non", "binaires", "donne", "la", "matrice", "de", "confusion", "."], "sentence-detokenized": "L'extension de ce concept aux classifications non binaires donne la matrice de confusion.", "token2charspan": [[0, 2], [2, 11], [12, 14], [15, 17], [18, 25], [26, 29], [30, 45], [46, 49], [50, 58], [59, 64], [65, 67], [68, 75], [76, 78], [79, 88], [88, 89]]}
{"doc_key": "ai-dev-67", "ner": [[17, 19, "algorithm"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Une", "estimation", "actualis\u00e9e", "de", "la", "variance", "du", "bruit", "de", "mesure", "peut", "\u00eatre", "obtenue", "par", "le", "calcul", "du", "maximum", "de", "vraisemblance", "."], "sentence-detokenized": "Une estimation actualis\u00e9e de la variance du bruit de mesure peut \u00eatre obtenue par le calcul du maximum de vraisemblance.", "token2charspan": [[0, 3], [4, 14], [15, 25], [26, 28], [29, 31], [32, 40], [41, 43], [44, 49], [50, 52], [53, 59], [60, 64], [65, 69], [70, 77], [78, 81], [82, 84], [85, 91], [92, 94], [95, 102], [103, 105], [106, 119], [119, 120]]}
{"doc_key": "ai-dev-68", "ner": [[1, 2, "field"], [5, 5, "algorithm"], [10, 11, "field"], [14, 15, "task"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[5, 5, 10, 11, "usage", "", true, false], [5, 5, 14, 15, "related-to", "", true, false], [10, 11, 1, 2, "type-of", "", true, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["En", "apprentissage", "automatique", ",", "le", "perceptron", "est", "un", "algorithme", "d'", "apprentissage", "supervis\u00e9", "de", "la", "classification", "binaire", "."], "sentence-detokenized": "En apprentissage automatique, le perceptron est un algorithme d'apprentissage supervis\u00e9 de la classification binaire.", "token2charspan": [[0, 2], [3, 16], [17, 28], [28, 29], [30, 32], [33, 43], [44, 47], [48, 50], [51, 61], [62, 64], [64, 77], [78, 87], [88, 90], [91, 93], [94, 108], [109, 116], [116, 117]]}
{"doc_key": "ai-dev-69", "ner": [[8, 9, "field"], [12, 12, "field"], [16, 25, "conference"], [28, 34, "conference"], [37, 46, "conference"], [49, 54, "conference"], [57, 62, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[16, 25, 8, 9, "topic", "", false, false], [16, 25, 12, 12, "topic", "", false, false], [28, 34, 8, 9, "topic", "", false, false], [28, 34, 12, 12, "topic", "", false, false], [37, 46, 8, 9, "topic", "", false, false], [37, 46, 12, 12, "topic", "", false, false], [49, 54, 8, 9, "topic", "", false, false], [49, 54, 12, 12, "topic", "", false, false], [57, 62, 8, 9, "topic", "", false, false], [57, 62, 12, 12, "topic", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9], "sentence": ["Elle", "a", "\u00e9galement", "pr\u00e9sid\u00e9", "plusieurs", "conf\u00e9rences", "sur", "l'", "apprentissage", "automatique", "et", "la", "vision", ",", "notamment", "la", "Conf\u00e9rence", "sur", "les", "syst\u00e8mes", "de", "traitement", "de", "l'", "information", "neuronale", ",", "la", "Conf\u00e9rence", "internationale", "sur", "les", "repr\u00e9sentations", "d'", "apprentissage", ",", "la", "Conf\u00e9rence", "sur", "la", "vision", "informatique", "et", "la", "reconnaissance", "des", "formes", ",", "la", "Conf\u00e9rence", "internationale", "sur", "la", "vision", "informatique", "et", "la", "Conf\u00e9rence", "europ\u00e9enne", "sur", "la", "vision", "informatique", "."], "sentence-detokenized": "Elle a \u00e9galement pr\u00e9sid\u00e9 plusieurs conf\u00e9rences sur l'apprentissage automatique et la vision, notamment la Conf\u00e9rence sur les syst\u00e8mes de traitement de l'information neuronale, la Conf\u00e9rence internationale sur les repr\u00e9sentations d'apprentissage, la Conf\u00e9rence sur la vision informatique et la reconnaissance des formes, la Conf\u00e9rence internationale sur la vision informatique et la Conf\u00e9rence europ\u00e9enne sur la vision informatique.", "token2charspan": [[0, 4], [5, 6], [7, 16], [17, 24], [25, 34], [35, 46], [47, 50], [51, 53], [53, 66], [67, 78], [79, 81], [82, 84], [85, 91], [91, 92], [93, 102], [103, 105], [106, 116], [117, 120], [121, 124], [125, 133], [134, 136], [137, 147], [148, 150], [151, 153], [153, 164], [165, 174], [174, 175], [176, 178], [179, 189], [190, 204], [205, 208], [209, 212], [213, 228], [229, 231], [231, 244], [244, 245], [246, 248], [249, 259], [260, 263], [264, 266], [267, 273], [274, 286], [287, 289], [290, 292], [293, 307], [308, 311], [312, 318], [318, 319], [320, 322], [323, 333], [334, 348], [349, 352], [353, 355], [356, 362], [363, 375], [376, 378], [379, 381], [382, 392], [393, 403], [404, 407], [408, 410], [411, 417], [418, 430], [430, 431]]}
{"doc_key": "ai-dev-70", "ner": [[1, 3, "algorithm"], [10, 13, "product"]], "ner_mapping_to_source": [0, 1], "relations": [[10, 13, 1, 3, "usage", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["L'", "algorithme", "de", "condensation", "a", "\u00e9galement", "\u00e9t\u00e9", "utilis\u00e9", "pour", "un", "syst\u00e8me", "de", "reconnaissance", "faciale", "dans", "une", "s\u00e9quence", "vid\u00e9o", "."], "sentence-detokenized": "L'algorithme de condensation a \u00e9galement \u00e9t\u00e9 utilis\u00e9 pour un syst\u00e8me de reconnaissance faciale dans une s\u00e9quence vid\u00e9o.", "token2charspan": [[0, 2], [2, 12], [13, 15], [16, 28], [29, 30], [31, 40], [41, 44], [45, 52], [53, 57], [58, 60], [61, 68], [69, 71], [72, 86], [87, 94], [95, 99], [100, 103], [104, 112], [113, 118], [118, 119]]}
{"doc_key": "ai-dev-71", "ner": [[0, 5, "task"], [12, 12, "organisation"], [27, 27, "conference"], [31, 35, "academicjournal"], [38, 38, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[27, 27, 0, 5, "topic", "", false, false], [27, 27, 12, 12, "origin", "", false, false], [31, 35, 0, 5, "topic", "", false, false], [31, 35, 12, 12, "origin", "", true, false], [38, 38, 31, 35, "role", "edited_by", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["La", "diffusion", "de", "l'", "information", "fait", "\u00e9galement", "partie", "des", "missions", "de", "l'", "ELRA", ",", "ce", "qui", "se", "traduit", "\u00e0", "la", "fois", "par", "l'", "organisation", "de", "la", "conf\u00e9rence", "LREC", "et", "par", "le", "Language", "Resources", "and", "Evaluation", "Journal", "\u00e9dit\u00e9", "par", "Springer", "."], "sentence-detokenized": "La diffusion de l'information fait \u00e9galement partie des missions de l'ELRA, ce qui se traduit \u00e0 la fois par l'organisation de la conf\u00e9rence LREC et par le Language Resources and Evaluation Journal \u00e9dit\u00e9 par Springer.", "token2charspan": [[0, 2], [3, 12], [13, 15], [16, 18], [18, 29], [30, 34], [35, 44], [45, 51], [52, 55], [56, 64], [65, 67], [68, 70], [70, 74], [74, 75], [76, 78], [79, 82], [83, 85], [86, 93], [94, 95], [96, 98], [99, 103], [104, 107], [108, 110], [110, 122], [123, 125], [126, 128], [129, 139], [140, 144], [145, 147], [148, 151], [152, 154], [155, 163], [164, 173], [174, 177], [178, 188], [189, 196], [197, 202], [203, 206], [207, 215], [215, 216]]}
{"doc_key": "ai-dev-72", "ner": [[2, 11, "field"], [15, 17, "field"], [20, 23, "field"], [26, 28, "field"], [66, 67, "field"], [74, 74, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[2, 11, 66, 67, "named", "", false, false], [20, 23, 2, 11, "named", "", false, false], [74, 74, 15, 17, "part-of", "", true, false], [74, 74, 20, 23, "part-of", "", true, false], [74, 74, 66, 67, "part-of", "", true, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Dans", "la", "th\u00e9orie", "des", "syst\u00e8mes", "lin\u00e9aires", "invariants", "dans", "le", "temps", "(", "LTI", ")", ",", "la", "th\u00e9orie", "du", "contr\u00f4le", "et", "le", "traitement", "des", "signaux", "num\u00e9riques", "ou", "le", "traitement", "des", "signaux", ",", "la", "relation", "entre", "le", "signal", "d'", "entr\u00e9e", ",", "math", "\\", "displaystyle", "x", "(", "t", ")", "/", "math", ",", "et", "le", "signal", "de", "sortie", ",", "math", "\\", "displaystyle", "y", "(", "t", ")", "/", "math", ",", "d'", "un", "syst\u00e8me", "LTI", "est", "r\u00e9gie", "par", "une", "op\u00e9ration", "de", "convolution", ":"], "sentence-detokenized": "Dans la th\u00e9orie des syst\u00e8mes lin\u00e9aires invariants dans le temps (LTI), la th\u00e9orie du contr\u00f4le et le traitement des signaux num\u00e9riques ou le traitement des signaux, la relation entre le signal d'entr\u00e9e, math\\ displaystyle x (t) / math, et le signal de sortie, math\\ displaystyle y (t) / math, d'un syst\u00e8me LTI est r\u00e9gie par une op\u00e9ration de convolution :", "token2charspan": [[0, 4], [5, 7], [8, 15], [16, 19], [20, 28], [29, 38], [39, 49], [50, 54], [55, 57], [58, 63], [64, 65], [65, 68], [68, 69], [69, 70], [71, 73], [74, 81], [82, 84], [85, 93], [94, 96], [97, 99], [100, 110], [111, 114], [115, 122], [123, 133], [134, 136], [137, 139], [140, 150], [151, 154], [155, 162], [162, 163], [164, 166], [167, 175], [176, 181], [182, 184], [185, 191], [192, 194], [194, 200], [200, 201], [202, 206], [206, 207], [208, 220], [221, 222], [223, 224], [224, 225], [225, 226], [227, 228], [229, 233], [233, 234], [235, 237], [238, 240], [241, 247], [248, 250], [251, 257], [257, 258], [259, 263], [263, 264], [265, 277], [278, 279], [280, 281], [281, 282], [282, 283], [284, 285], [286, 290], [290, 291], [292, 294], [294, 296], [297, 304], [305, 308], [309, 312], [313, 318], [319, 322], [323, 326], [327, 336], [337, 339], [340, 351], [352, 353]]}
{"doc_key": "ai-dev-73", "ner": [[19, 21, "field"], [24, 26, "field"], [29, 30, "field"], [33, 36, "field"], [39, 43, "field"], [46, 47, "product"], [50, 51, "field"], [54, 54, "field"], [57, 58, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "relations": [], "relations_mapping_to_source": [], "sentence": ["En", "raison", "de", "sa", "g\u00e9n\u00e9ralit\u00e9", ",", "le", "domaine", "est", "\u00e9tudi\u00e9", "dans", "de", "nombreuses", "autres", "disciplines", ",", "telles", "que", "la", "th\u00e9orie", "des", "jeux", ",", "la", "th\u00e9orie", "du", "contr\u00f4le", ",", "la", "recherche", "op\u00e9rationnelle", ",", "la", "th\u00e9orie", "de", "l'", "information", ",", "l'", "optimisation", "bas\u00e9e", "sur", "la", "simulation", ",", "les", "syst\u00e8mes", "multi-agents", ",", "l'", "intelligence", "artificielle", ",", "les", "statistiques", "et", "les", "algorithmes", "g\u00e9n\u00e9tiques", "."], "sentence-detokenized": "En raison de sa g\u00e9n\u00e9ralit\u00e9, le domaine est \u00e9tudi\u00e9 dans de nombreuses autres disciplines, telles que la th\u00e9orie des jeux, la th\u00e9orie du contr\u00f4le, la recherche op\u00e9rationnelle, la th\u00e9orie de l'information, l'optimisation bas\u00e9e sur la simulation, les syst\u00e8mes multi-agents, l'intelligence artificielle, les statistiques et les algorithmes g\u00e9n\u00e9tiques.", "token2charspan": [[0, 2], [3, 9], [10, 12], [13, 15], [16, 26], [26, 27], [28, 30], [31, 38], [39, 42], [43, 49], [50, 54], [55, 57], [58, 68], [69, 75], [76, 87], [87, 88], [89, 95], [96, 99], [100, 102], [103, 110], [111, 114], [115, 119], [119, 120], [121, 123], [124, 131], [132, 134], [135, 143], [143, 144], [145, 147], [148, 157], [158, 172], [172, 173], [174, 176], [177, 184], [185, 187], [188, 190], [190, 201], [201, 202], [203, 205], [205, 217], [218, 223], [224, 227], [228, 230], [231, 241], [241, 242], [243, 246], [247, 255], [256, 268], [268, 269], [270, 272], [272, 284], [285, 297], [297, 298], [299, 302], [303, 315], [316, 318], [319, 322], [323, 334], [335, 345], [345, 346]]}
{"doc_key": "ai-dev-74", "ner": [[0, 4, "algorithm"], [19, 20, "field"], [26, 29, "algorithm"], [35, 36, "algorithm"], [43, 44, "algorithm"], [48, 49, "algorithm"], [48, 51, "researcher"], [53, 54, "researcher"], [56, 58, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "relations": [[19, 20, 0, 4, "usage", "", true, false], [26, 29, 19, 20, "part-of", "", true, false], [35, 36, 19, 20, "part-of", "", true, false], [43, 44, 19, 20, "part-of", "", true, false], [48, 49, 19, 20, "part-of", "", true, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["La", "descente", "de", "gradient", "stochastique", "est", "un", "algorithme", "populaire", "pour", "l'", "entra\u00eenement", "d'", "un", "large", "\u00e9ventail", "de", "mod\u00e8les", "en", "apprentissage", "automatique", ",", "y", "compris", "les", "machines", "\u00e0", "vecteurs", "de", "support", "(", "lin\u00e9aires", ")", ",", "la", "r\u00e9gression", "logistique", "(", "voir", ",", "par", "exemple", ",", "Vowpal", "Wabbit", ")", "et", "les", "mod\u00e8les", "graphiques.Jenny", "Rose", "Finkel", ",", "Alex", "Kleeman", ",", "Christopher", "D.", "Manning", "(", "2008", ")", "."], "sentence-detokenized": "La descente de gradient stochastique est un algorithme populaire pour l'entra\u00eenement d'un large \u00e9ventail de mod\u00e8les en apprentissage automatique, y compris les machines \u00e0 vecteurs de support (lin\u00e9aires), la r\u00e9gression logistique (voir, par exemple, Vowpal Wabbit) et les mod\u00e8les graphiques.Jenny Rose Finkel, Alex Kleeman, Christopher D. Manning (2008).", "token2charspan": [[0, 2], [3, 11], [12, 14], [15, 23], [24, 36], [37, 40], [41, 43], [44, 54], [55, 64], [65, 69], [70, 72], [72, 84], [85, 87], [87, 89], [90, 95], [96, 104], [105, 107], [108, 115], [116, 118], [119, 132], [133, 144], [144, 145], [146, 147], [148, 155], [156, 159], [160, 168], [169, 170], [171, 179], [180, 182], [183, 190], [191, 192], [192, 201], [201, 202], [202, 203], [204, 206], [207, 217], [218, 228], [229, 230], [230, 234], [234, 235], [236, 239], [240, 247], [247, 248], [249, 255], [256, 262], [262, 263], [264, 266], [267, 270], [271, 278], [279, 295], [296, 300], [301, 307], [307, 308], [309, 313], [314, 321], [321, 322], [323, 334], [335, 337], [338, 345], [346, 347], [347, 351], [351, 352], [352, 353]]}
{"doc_key": "ai-dev-75", "ner": [[9, 9, "organisation"], [14, 15, "product"], [22, 22, "country"], [25, 28, "university"], [30, 30, "location"], [33, 35, "university"], [37, 37, "location"], [40, 41, "university"], [43, 43, "location"], [46, 48, "university"], [50, 50, "location"], [53, 54, "university"], [56, 56, "location"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12], "relations": [[9, 9, 25, 28, "role", "donates_to", false, false], [9, 9, 33, 35, "role", "donates_to", false, false], [9, 9, 40, 41, "role", "donates_to", false, false], [9, 9, 46, 48, "role", "donates_to", false, false], [9, 9, 53, 54, "role", "donates_to", false, false], [14, 15, 9, 9, "origin", "donates", true, false], [25, 28, 30, 30, "physical", "", false, false], [30, 30, 22, 22, "physical", "", false, false], [33, 35, 37, 37, "physical", "", false, false], [37, 37, 22, 22, "physical", "", false, false], [40, 41, 43, 43, "physical", "", false, false], [43, 43, 22, 22, "physical", "", false, false], [46, 48, 50, 50, "physical", "", false, false], [50, 50, 22, 22, "physical", "", false, false], [53, 54, 56, 56, "physical", "", false, false], [56, 56, 22, 22, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15], "sentence": ["En", "ao\u00fbt", "2011", ",", "il", "a", "\u00e9t\u00e9", "annonc\u00e9", "qu'", "Hitachi", "ferait", "don", "d'", "un", "microscope", "\u00e9lectronique", "\u00e0", "chacune", "des", "cinq", "universit\u00e9s", "d'", "Indon\u00e9sie", "(", "l'", "universit\u00e9", "de", "Sumatra", "Nord", "\u00e0", "Medan", ",", "l'", "universit\u00e9", "chr\u00e9tienne", "indon\u00e9sienne", "\u00e0", "Jakarta", ",", "l'", "universit\u00e9", "Padjadjaran", "\u00e0", "Bandung", ",", "l'", "universit\u00e9", "Jenderal", "Soedirman", "\u00e0", "Purwokerto", "et", "l'", "universit\u00e9", "Muhammadiyah", "\u00e0", "Malang", ")", "."], "sentence-detokenized": "En ao\u00fbt 2011, il a \u00e9t\u00e9 annonc\u00e9 qu'Hitachi ferait don d'un microscope \u00e9lectronique \u00e0 chacune des cinq universit\u00e9s d'Indon\u00e9sie (l'universit\u00e9 de Sumatra Nord \u00e0 Medan, l'universit\u00e9 chr\u00e9tienne indon\u00e9sienne \u00e0 Jakarta, l'universit\u00e9 Padjadjaran \u00e0 Bandung, l'universit\u00e9 Jenderal Soedirman \u00e0 Purwokerto et l'universit\u00e9 Muhammadiyah \u00e0 Malang).", "token2charspan": [[0, 2], [3, 7], [8, 12], [12, 13], [14, 16], [17, 18], [19, 22], [23, 30], [31, 34], [34, 41], [42, 48], [49, 52], [53, 55], [55, 57], [58, 68], [69, 81], [82, 83], [84, 91], [92, 95], [96, 100], [101, 112], [113, 115], [115, 124], [125, 126], [126, 128], [128, 138], [139, 141], [142, 149], [150, 154], [155, 156], [157, 162], [162, 163], [164, 166], [166, 176], [177, 187], [188, 200], [201, 202], [203, 210], [210, 211], [212, 214], [214, 224], [225, 236], [237, 238], [239, 246], [246, 247], [248, 250], [250, 260], [261, 269], [270, 279], [280, 281], [282, 292], [293, 295], [296, 298], [298, 308], [309, 321], [322, 323], [324, 330], [330, 331], [331, 332]]}
{"doc_key": "ai-dev-76", "ner": [[3, 3, "field"], [6, 7, "field"], [12, 13, "algorithm"], [16, 17, "algorithm"], [27, 28, "field"], [36, 37, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[3, 3, 6, 7, "part-of", "", false, false], [3, 3, 27, 28, "related-to", "", true, false], [3, 3, 36, 37, "related-to", "", true, false], [12, 13, 3, 3, "type-of", "", false, false], [16, 17, 3, 3, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Les", "techniques", "d'", "optimisation", "de", "la", "recherche", "op\u00e9rationnelle", ",", "telles", "que", "la", "programmation", "lin\u00e9aire", "ou", "la", "programmation", "dynamique", ",", "sont", "souvent", "peu", "pratiques", "pour", "les", "probl\u00e8mes", "de", "g\u00e9nie", "logiciel", "\u00e0", "grande", "\u00e9chelle", "en", "raison", "de", "leur", "complexit\u00e9", "informatique", "."], "sentence-detokenized": "Les techniques d'optimisation de la recherche op\u00e9rationnelle, telles que la programmation lin\u00e9aire ou la programmation dynamique, sont souvent peu pratiques pour les probl\u00e8mes de g\u00e9nie logiciel \u00e0 grande \u00e9chelle en raison de leur complexit\u00e9 informatique.", "token2charspan": [[0, 3], [4, 14], [15, 17], [17, 29], [30, 32], [33, 35], [36, 45], [46, 60], [60, 61], [62, 68], [69, 72], [73, 75], [76, 89], [90, 98], [99, 101], [102, 104], [105, 118], [119, 128], [128, 129], [130, 134], [135, 142], [143, 146], [147, 156], [157, 161], [162, 165], [166, 175], [176, 178], [179, 184], [185, 193], [194, 195], [196, 202], [203, 210], [211, 213], [214, 220], [221, 223], [224, 228], [229, 239], [240, 252], [252, 253]]}
{"doc_key": "ai-dev-77", "ner": [[1, 1, "metrics"], [6, 6, "metrics"], [10, 12, "metrics"], [17, 18, "metrics"], [21, 24, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[1, 1, 6, 6, "compare", "", false, false], [1, 1, 10, 12, "compare", "", false, false], [17, 18, 10, 12, "part-of", "", false, false], [21, 24, 10, 12, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["La", "sensibilit\u00e9", "est", "diff\u00e9rente", "de", "la", "pr\u00e9cision", "ou", "de", "la", "valeur", "pr\u00e9dictive", "positive", "(", "rapport", "entre", "les", "VRAIS", "positifs", "et", "les", "VRAIS", "et", "FAUX", "positifs", "combin\u00e9s", ")", ",", "qui", "est", "autant", "une", "d\u00e9claration", "sur", "la", "proportion", "de", "positifs", "r\u00e9els", "dans", "la", "population", "test\u00e9e", "que", "sur", "le", "test", "."], "sentence-detokenized": "La sensibilit\u00e9 est diff\u00e9rente de la pr\u00e9cision ou de la valeur pr\u00e9dictive positive (rapport entre les VRAIS positifs et les VRAIS et FAUX positifs combin\u00e9s), qui est autant une d\u00e9claration sur la proportion de positifs r\u00e9els dans la population test\u00e9e que sur le test.", "token2charspan": [[0, 2], [3, 14], [15, 18], [19, 29], [30, 32], [33, 35], [36, 45], [46, 48], [49, 51], [52, 54], [55, 61], [62, 72], [73, 81], [82, 83], [83, 90], [91, 96], [97, 100], [101, 106], [107, 115], [116, 118], [119, 122], [123, 128], [129, 131], [132, 136], [137, 145], [146, 154], [154, 155], [155, 156], [157, 160], [161, 164], [165, 171], [172, 175], [176, 187], [188, 191], [192, 194], [195, 205], [206, 208], [209, 217], [218, 223], [224, 228], [229, 231], [232, 242], [243, 249], [250, 253], [254, 257], [258, 260], [261, 265], [265, 266]]}
{"doc_key": "ai-dev-78", "ner": [[3, 4, "person"], [13, 13, "product"], [18, 18, "person"], [38, 38, "person"], [44, 45, "person"], [52, 52, "person"], [58, 59, "person"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[3, 4, 52, 52, "named", "same", false, false], [13, 13, 3, 4, "artifact", "", false, false], [44, 45, 58, 59, "role", "convinces", false, false], [58, 59, 13, 13, "role", "producer", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Le", "sc\u00e9nario", "d'", "Hampton", "Fancher", "!", "-", "-", "qui", "ne", "s'", "intitulait", "pas", "Android", "au", "d\u00e9part", "-", "voir", "Sammon", ",", "pp.", "32", "et", "38", "pour", "l'", "explication", "--", "a", "fait", "l'", "objet", "d'", "une", "option", "en", "1977", ".", "Sammon", ",", "pp.", "23-30", "Le", "producteur", "Michael", "Deeley", "s'", "est", "int\u00e9ress\u00e9", "au", "projet", "de", "Fancher", "et", "a", "convaincu", "le", "r\u00e9alisateur", "Ridley", "Scott", "de", "le", "tourner", "."], "sentence-detokenized": "Le sc\u00e9nario d'Hampton Fancher ! -- qui ne s'intitulait pas Android au d\u00e9part - voir Sammon, pp. 32 et 38 pour l'explication -- a fait l'objet d'une option en 1977. Sammon, pp. 23-30 Le producteur Michael Deeley s'est int\u00e9ress\u00e9 au projet de Fancher et a convaincu le r\u00e9alisateur Ridley Scott de le tourner.", "token2charspan": [[0, 2], [3, 11], [12, 14], [14, 21], [22, 29], [30, 31], [32, 33], [33, 34], [35, 38], [39, 41], [42, 44], [44, 54], [55, 58], [59, 66], [67, 69], [70, 76], [77, 78], [79, 83], [84, 90], [90, 91], [92, 95], [96, 98], [99, 101], [102, 104], [105, 109], [110, 112], [112, 123], [124, 126], [127, 128], [129, 133], [134, 136], [136, 141], [142, 144], [144, 147], [148, 154], [155, 157], [158, 162], [162, 163], [164, 170], [170, 171], [172, 175], [176, 181], [182, 184], [185, 195], [196, 203], [204, 210], [211, 213], [213, 216], [217, 226], [227, 229], [230, 236], [237, 239], [240, 247], [248, 250], [251, 252], [253, 262], [263, 265], [266, 277], [278, 284], [285, 290], [291, 293], [294, 296], [297, 304], [304, 305]]}
{"doc_key": "ai-dev-79", "ner": [[0, 3, "field"], [6, 8, "task"], [11, 12, "task"], [16, 21, "misc"], [24, 26, "field"], [29, 32, "task"], [35, 37, "task"], [42, 44, "field"], [48, 53, "task"], [56, 56, "task"], [59, 60, "task"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10], "relations": [[6, 8, 0, 3, "part-of", "", false, false], [11, 12, 0, 3, "part-of", "", false, false], [16, 21, 0, 3, "part-of", "", false, false], [24, 26, 0, 3, "part-of", "", false, false], [29, 32, 0, 3, "part-of", "", false, false], [35, 37, 0, 3, "part-of", "", false, false], [42, 44, 0, 3, "part-of", "", false, false], [48, 53, 0, 3, "part-of", "", false, false], [56, 56, 0, 3, "part-of", "", false, false], [59, 60, 0, 3, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9], "sentence": ["L'", "analyse", "de", "texte", "comprend", "la", "recherche", "d'", "informations", ",", "l'", "analyse", "lexicale", "pour", "\u00e9tudier", "la", "distribution", "de", "la", "fr\u00e9quence", "des", "mots", ",", "la", "reconnaissance", "des", "formes", ",", "l'", "\u00e9tiquetage", "et", "l'", "annotation", ",", "l'", "extraction", "d'", "informations", ",", "les", "techniques", "d'", "exploration", "de", "donn\u00e9es", ",", "notamment", "l'", "analyse", "des", "liens", "et", "des", "associations", ",", "la", "visualisation", "et", "l'", "analyse", "pr\u00e9dictive", "."], "sentence-detokenized": "L'analyse de texte comprend la recherche d'informations, l'analyse lexicale pour \u00e9tudier la distribution de la fr\u00e9quence des mots, la reconnaissance des formes, l'\u00e9tiquetage et l'annotation, l'extraction d'informations, les techniques d'exploration de donn\u00e9es, notamment l'analyse des liens et des associations, la visualisation et l'analyse pr\u00e9dictive.", "token2charspan": [[0, 2], [2, 9], [10, 12], [13, 18], [19, 27], [28, 30], [31, 40], [41, 43], [43, 55], [55, 56], [57, 59], [59, 66], [67, 75], [76, 80], [81, 88], [89, 91], [92, 104], [105, 107], [108, 110], [111, 120], [121, 124], [125, 129], [129, 130], [131, 133], [134, 148], [149, 152], [153, 159], [159, 160], [161, 163], [163, 173], [174, 176], [177, 179], [179, 189], [189, 190], [191, 193], [193, 203], [204, 206], [206, 218], [218, 219], [220, 223], [224, 234], [235, 237], [237, 248], [249, 251], [252, 259], [259, 260], [261, 270], [271, 273], [273, 280], [281, 284], [285, 290], [291, 293], [294, 297], [298, 310], [310, 311], [312, 314], [315, 328], [329, 331], [332, 334], [334, 341], [342, 352], [352, 353]]}
{"doc_key": "ai-dev-80", "ner": [[3, 3, "product"], [11, 12, "misc"]], "ner_mapping_to_source": [0, 1], "relations": [[11, 12, 3, 3, "part-of", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Plusieurs", "mesures", "utilisent", "WordNet", ",", "une", "base", "de", "donn\u00e9es", "lexicale", "de", "mots", "anglais", "construite", "manuellement", "."], "sentence-detokenized": "Plusieurs mesures utilisent WordNet, une base de donn\u00e9es lexicale de mots anglais construite manuellement.", "token2charspan": [[0, 9], [10, 17], [18, 27], [28, 35], [35, 36], [37, 40], [41, 45], [46, 48], [49, 56], [57, 65], [66, 68], [69, 73], [74, 81], [82, 92], [93, 105], [105, 106]]}
{"doc_key": "ai-dev-81", "ner": [[8, 9, "field"], [12, 14, "task"], [17, 23, "task"]], "ner_mapping_to_source": [0, 1, 2], "relations": [], "relations_mapping_to_source": [], "sentence": ["Le", "syst\u00e8me", "utilise", "une", "combinaison", "de", "techniques", "de", "linguistique", "informatique", ",", "de", "recherche", "d'", "informations", "et", "de", "repr\u00e9sentation", "des", "connaissances", "pour", "trouver", "des", "r\u00e9ponses", "."], "sentence-detokenized": "Le syst\u00e8me utilise une combinaison de techniques de linguistique informatique, de recherche d'informations et de repr\u00e9sentation des connaissances pour trouver des r\u00e9ponses.", "token2charspan": [[0, 2], [3, 10], [11, 18], [19, 22], [23, 34], [35, 37], [38, 48], [49, 51], [52, 64], [65, 77], [77, 78], [79, 81], [82, 91], [92, 94], [94, 106], [107, 109], [110, 112], [113, 127], [128, 131], [132, 145], [146, 150], [151, 158], [159, 162], [163, 171], [171, 172]]}
{"doc_key": "ai-dev-82", "ner": [[8, 10, "metrics"], [20, 22, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[8, 10, 20, 22, "compare", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["En", "tant", "que", "mesure", "de", "performance", ",", "le", "coefficient", "d'", "incertitude", "pr\u00e9sente", "l'", "avantage", ",", "par", "rapport", "\u00e0", "la", "simple", "pr\u00e9cision", ",", "de", "ne", "pas", "\u00eatre", "affect\u00e9", "par", "les", "tailles", "relatives", "des", "diff\u00e9rentes", "classes", "."], "sentence-detokenized": "En tant que mesure de performance, le coefficient d'incertitude pr\u00e9sente l'avantage, par rapport \u00e0 la simple pr\u00e9cision, de ne pas \u00eatre affect\u00e9 par les tailles relatives des diff\u00e9rentes classes.", "token2charspan": [[0, 2], [3, 7], [8, 11], [12, 18], [19, 21], [22, 33], [33, 34], [35, 37], [38, 49], [50, 52], [52, 63], [64, 72], [73, 75], [75, 83], [83, 84], [85, 88], [89, 96], [97, 98], [99, 101], [102, 108], [109, 118], [118, 119], [120, 122], [123, 125], [126, 129], [130, 134], [135, 142], [143, 146], [147, 150], [151, 158], [159, 168], [169, 172], [173, 184], [185, 192], [192, 193]]}
{"doc_key": "ai-dev-83", "ner": [[12, 13, "algorithm"], [16, 18, "algorithm"], [22, 24, "algorithm"]], "ner_mapping_to_source": [0, 1, 2], "relations": [], "relations_mapping_to_source": [], "sentence": ["Les", "chercheurs", "ont", "essay\u00e9", "un", "certain", "nombre", "de", "m\u00e9thodes", "telles", "que", "le", "flux", "optique", ",", "le", "filtrage", "de", "Kalman", ",", "les", "mod\u00e8les", "de", "Markov", "cach\u00e9s", ",", "etc."], "sentence-detokenized": "Les chercheurs ont essay\u00e9 un certain nombre de m\u00e9thodes telles que le flux optique, le filtrage de Kalman, les mod\u00e8les de Markov cach\u00e9s, etc.", "token2charspan": [[0, 3], [4, 14], [15, 18], [19, 25], [26, 28], [29, 36], [37, 43], [44, 46], [47, 55], [56, 62], [63, 66], [67, 69], [70, 74], [75, 82], [82, 83], [84, 86], [87, 95], [96, 98], [99, 105], [105, 106], [107, 110], [111, 118], [119, 121], [122, 128], [129, 135], [135, 136], [137, 141]]}
{"doc_key": "ai-dev-84", "ner": [[13, 16, "conference"], [31, 33, "organisation"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Elle", "a", "occup\u00e9", "les", "postes", "de", "pr\u00e9sidente", ",", "vice-pr\u00e9sidente", "et", "secr\u00e9taire-tr\u00e9sori\u00e8re", "de", "l'", "Association", "for", "Computational", "Linguistics", "et", "a", "\u00e9t\u00e9", "membre", "du", "conseil", "d'", "administration", "et", "secr\u00e9taire", "du", "conseil", "de", "la", "Computing", "Research", "Association", "."], "sentence-detokenized": "Elle a occup\u00e9 les postes de pr\u00e9sidente, vice-pr\u00e9sidente et secr\u00e9taire-tr\u00e9sori\u00e8re de l'Association for Computational Linguistics et a \u00e9t\u00e9 membre du conseil d'administration et secr\u00e9taire du conseil de la Computing Research Association.", "token2charspan": [[0, 4], [5, 6], [7, 13], [14, 17], [18, 24], [25, 27], [28, 38], [38, 39], [40, 55], [56, 58], [59, 80], [81, 83], [84, 86], [86, 97], [98, 101], [102, 115], [116, 127], [128, 130], [131, 132], [133, 136], [137, 143], [144, 146], [147, 154], [155, 157], [157, 171], [172, 174], [175, 185], [186, 188], [189, 196], [197, 199], [200, 202], [203, 212], [213, 221], [222, 233], [233, 234]]}
{"doc_key": "ai-dev-85", "ner": [[7, 7, "programlang"], [9, 9, "product"], [11, 11, "programlang"], [14, 15, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[7, 7, 11, 11, "compare", "", false, false], [7, 7, 14, 15, "related-to", "supports", false, false], [9, 9, 11, 11, "compare", "", false, false], [9, 9, 14, 15, "related-to", "supports", false, false], [11, 11, 14, 15, "related-to", "supports", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Comme", "d'", "autres", "langages", "similaires", "tels", "que", "APL", "et", "MATLAB", ",", "R", "supporte", "l'", "arithm\u00e9tique", "matricielle", "."], "sentence-detokenized": "Comme d'autres langages similaires tels que APL et MATLAB, R supporte l'arithm\u00e9tique matricielle.", "token2charspan": [[0, 5], [6, 8], [8, 14], [15, 23], [24, 34], [35, 39], [40, 43], [44, 47], [48, 50], [51, 57], [57, 58], [59, 60], [61, 69], [70, 72], [72, 84], [85, 96], [96, 97]]}
{"doc_key": "ai-dev-86", "ner": [[10, 12, "misc"], [15, 16, "organisation"], [20, 21, "researcher"], [24, 26, "university"], [30, 36, "misc"], [38, 38, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[10, 12, 15, 16, "physical", "", false, false], [10, 12, 30, 36, "temporal", "", false, false], [20, 21, 10, 12, "role", "arranges", false, false], [20, 21, 24, 26, "role", "works_for", false, false], [38, 38, 10, 12, "win-defeat", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Le", "7", "juin", "2014", ",", "lors", "d'", "un", "concours", "de", "tests", "de", "Turing", "\u00e0", "la", "Royal", "Society", ",", "organis\u00e9", "par", "Kevin", "Warwick", "de", "l'", "universit\u00e9", "de", "Reading", "pour", "marquer", "le", "60e", "anniversaire", "de", "la", "mort", "de", "Turing", ",", "Goostman", "a", "gagn\u00e9", "apr\u00e8s", "que", "33", "%", "des", "juges", "aient", "\u00e9t\u00e9", "convaincus", "que", "le", "bot", "\u00e9tait", "humain", "."], "sentence-detokenized": "Le 7 juin 2014, lors d'un concours de tests de Turing \u00e0 la Royal Society, organis\u00e9 par Kevin Warwick de l'universit\u00e9 de Reading pour marquer le 60e anniversaire de la mort de Turing, Goostman a gagn\u00e9 apr\u00e8s que 33 % des juges aient \u00e9t\u00e9 convaincus que le bot \u00e9tait humain.", "token2charspan": [[0, 2], [3, 4], [5, 9], [10, 14], [14, 15], [16, 20], [21, 23], [23, 25], [26, 34], [35, 37], [38, 43], [44, 46], [47, 53], [54, 55], [56, 58], [59, 64], [65, 72], [72, 73], [74, 82], [83, 86], [87, 92], [93, 100], [101, 103], [104, 106], [106, 116], [117, 119], [120, 127], [128, 132], [133, 140], [141, 143], [144, 147], [148, 160], [161, 163], [164, 166], [167, 171], [172, 174], [175, 181], [181, 182], [183, 191], [192, 193], [194, 199], [200, 205], [206, 209], [210, 212], [213, 214], [215, 218], [219, 224], [225, 230], [231, 234], [235, 245], [246, 249], [250, 252], [253, 256], [257, 262], [263, 269], [269, 270]]}
{"doc_key": "ai-dev-87", "ner": [[1, 2, "product"], [4, 4, "product"]], "ner_mapping_to_source": [0, 1], "relations": [[4, 4, 1, 2, "named", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Un", "robot", "collaboratif", "ou", "cobot", "est", "un", "robot", "qui", "peut", "interagir", "de", "mani\u00e8re", "s\u00fbre", "et", "efficace", "avec", "des", "travailleurs", "humains", "tout", "en", "effectuant", "des", "t\u00e2ches", "industrielles", "simples", "."], "sentence-detokenized": "Un robot collaboratif ou cobot est un robot qui peut interagir de mani\u00e8re s\u00fbre et efficace avec des travailleurs humains tout en effectuant des t\u00e2ches industrielles simples.", "token2charspan": [[0, 2], [3, 8], [9, 21], [22, 24], [25, 30], [31, 34], [35, 37], [38, 43], [44, 47], [48, 52], [53, 62], [63, 65], [66, 73], [74, 78], [79, 81], [82, 90], [91, 95], [96, 99], [100, 112], [113, 120], [121, 125], [126, 128], [129, 139], [140, 143], [144, 150], [151, 164], [165, 172], [172, 173]]}
{"doc_key": "ai-dev-88", "ner": [[13, 15, "field"], [19, 21, "task"], [24, 26, "task"], [29, 31, "task"], [34, 36, "task"], [39, 41, "task"], [44, 48, "task"], [51, 53, "task"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "relations": [[19, 21, 13, 15, "part-of", "task_part_of_field", false, false], [24, 26, 13, 15, "part-of", "task_part_of_field", false, false], [29, 31, 13, 15, "part-of", "task_part_of_field", false, false], [34, 36, 13, 15, "part-of", "task_part_of_field", false, false], [39, 41, 13, 15, "part-of", "task_part_of_field", false, false], [44, 48, 13, 15, "part-of", "task_part_of_field", false, false], [51, 53, 13, 15, "part-of", "task_part_of_field", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "sentence": ["Ce", "cadre", "g\u00e9n\u00e9ral", "a", "\u00e9t\u00e9", "appliqu\u00e9", "\u00e0", "une", "grande", "vari\u00e9t\u00e9", "de", "probl\u00e8mes", "de", "vision", "par", "ordinateur", ",", "notamment", "la", "d\u00e9tection", "de", "caract\u00e9ristiques", ",", "la", "classification", "de", "caract\u00e9ristiques", ",", "la", "segmentation", "d'", "images", ",", "la", "correspondance", "d'", "images", ",", "l'", "estimation", "du", "mouvement", ",", "le", "calcul", "d'", "indices", "de", "forme", "et", "la", "reconnaissance", "d'", "objets", "."], "sentence-detokenized": "Ce cadre g\u00e9n\u00e9ral a \u00e9t\u00e9 appliqu\u00e9 \u00e0 une grande vari\u00e9t\u00e9 de probl\u00e8mes de vision par ordinateur, notamment la d\u00e9tection de caract\u00e9ristiques, la classification de caract\u00e9ristiques, la segmentation d'images, la correspondance d'images, l'estimation du mouvement, le calcul d'indices de forme et la reconnaissance d'objets.", "token2charspan": [[0, 2], [3, 8], [9, 16], [17, 18], [19, 22], [23, 31], [32, 33], [34, 37], [38, 44], [45, 52], [53, 55], [56, 65], [66, 68], [69, 75], [76, 79], [80, 90], [90, 91], [92, 101], [102, 104], [105, 114], [115, 117], [118, 134], [134, 135], [136, 138], [139, 153], [154, 156], [157, 173], [173, 174], [175, 177], [178, 190], [191, 193], [193, 199], [199, 200], [201, 203], [204, 218], [219, 221], [221, 227], [227, 228], [229, 231], [231, 241], [242, 244], [245, 254], [254, 255], [256, 258], [259, 265], [266, 268], [268, 275], [276, 278], [279, 284], [285, 287], [288, 290], [291, 305], [306, 308], [308, 314], [314, 315]]}
{"doc_key": "ai-dev-89", "ner": [[7, 9, "task"], [11, 14, "algorithm"], [19, 21, "algorithm"], [35, 36, "algorithm"], [40, 41, "algorithm"], [45, 46, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[7, 9, 11, 14, "part-of", "", false, false], [7, 9, 19, 21, "usage", "", false, false], [11, 14, 35, 36, "named", "same", false, false], [35, 36, 40, 41, "related-to", "", false, false], [35, 36, 45, 46, "related-to", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Dans", "de", "nombreuses", "applications", "pratiques", ",", "l'", "estimation", "des", "param\u00e8tres", "des", "mod\u00e8les", "de", "Bayes", "na\u00effs", "utilise", "la", "m\u00e9thode", "du", "maximum", "de", "vraisemblance", ";", "en", "d'", "autres", "termes", ",", "on", "peut", "travailler", "avec", "le", "mod\u00e8le", "de", "Bayes", "na\u00eff", "sans", "accepter", "la", "probabilit\u00e9", "bay\u00e9sienne", "ni", "utiliser", "de", "m\u00e9thodes", "bay\u00e9siennes", "."], "sentence-detokenized": "Dans de nombreuses applications pratiques, l'estimation des param\u00e8tres des mod\u00e8les de Bayes na\u00effs utilise la m\u00e9thode du maximum de vraisemblance ; en d'autres termes, on peut travailler avec le mod\u00e8le de Bayes na\u00eff sans accepter la probabilit\u00e9 bay\u00e9sienne ni utiliser de m\u00e9thodes bay\u00e9siennes.", "token2charspan": [[0, 4], [5, 7], [8, 18], [19, 31], [32, 41], [41, 42], [43, 45], [45, 55], [56, 59], [60, 70], [71, 74], [75, 82], [83, 85], [86, 91], [92, 97], [98, 105], [106, 108], [109, 116], [117, 119], [120, 127], [128, 130], [131, 144], [145, 146], [147, 149], [150, 152], [152, 158], [159, 165], [165, 166], [167, 169], [170, 174], [175, 185], [186, 190], [191, 193], [194, 200], [201, 203], [204, 209], [210, 214], [215, 219], [220, 228], [229, 231], [232, 243], [244, 254], [255, 257], [258, 266], [267, 269], [270, 278], [279, 290], [290, 291]]}
{"doc_key": "ai-dev-90", "ner": [[2, 4, "researcher"], [6, 7, "misc"], [11, 14, "university"], [16, 18, "researcher"], [20, 21, "misc"], [26, 26, "university"], [28, 28, "university"], [30, 30, "misc"], [38, 39, "university"], [45, 48, "misc"], [50, 53, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10], "relations": [[2, 4, 11, 14, "physical", "", false, false], [2, 4, 11, 14, "role", "", false, false], [2, 4, 16, 18, "social", "brothers", false, false], [6, 7, 2, 4, "named", "", false, false], [16, 18, 26, 26, "physical", "", false, false], [16, 18, 26, 26, "role", "", false, false], [16, 18, 28, 28, "physical", "", false, false], [16, 18, 28, 28, "role", "", false, false], [16, 18, 38, 39, "physical", "", false, false], [16, 18, 38, 39, "role", "", false, false], [20, 21, 16, 18, "named", "", false, false], [30, 30, 16, 18, "origin", "", false, false], [45, 48, 16, 18, "artifact", "", false, false], [45, 48, 50, 53, "part-of", "part", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13], "sentence": ["Fr\u00e8res", "-", "Victor", "Gershevich", "Katz", ",", "math\u00e9maticien", "am\u00e9ricain", ",", "professeur", "au", "Massachusetts", "Institute", "of", "Technology", ";", "Mikhail", "Gershevich", "Katz", ",", "math\u00e9maticien", "isra\u00e9lien", ",", "dipl\u00f4m\u00e9", "des", "universit\u00e9s", "Harvard", "et", "Columbia", "(", "Ph.D.", ",", "1984", ")", ",", "professeur", "\u00e0", "l'", "universit\u00e9", "Bar-Ilan", ",", "auteur", "de", "la", "monographie", "Systolic", "Geometry", "and", "Topology", "(", "Mathematical", "Surveys", "and", "Monographs", ",", "vol", "."], "sentence-detokenized": "Fr\u00e8res - Victor Gershevich Katz, math\u00e9maticien am\u00e9ricain, professeur au Massachusetts Institute of Technology ; Mikhail Gershevich Katz, math\u00e9maticien isra\u00e9lien, dipl\u00f4m\u00e9 des universit\u00e9s Harvard et Columbia (Ph.D., 1984), professeur \u00e0 l'universit\u00e9 Bar-Ilan, auteur de la monographie Systolic Geometry and Topology (Mathematical Surveys and Monographs, vol.", "token2charspan": [[0, 6], [7, 8], [9, 15], [16, 26], [27, 31], [31, 32], [33, 46], [47, 56], [56, 57], [58, 68], [69, 71], [72, 85], [86, 95], [96, 98], [99, 109], [110, 111], [112, 119], [120, 130], [131, 135], [135, 136], [137, 150], [151, 160], [160, 161], [162, 169], [170, 173], [174, 185], [186, 193], [194, 196], [197, 205], [206, 207], [207, 212], [212, 213], [214, 218], [218, 219], [219, 220], [221, 231], [232, 233], [234, 236], [236, 246], [247, 255], [255, 256], [257, 263], [264, 266], [267, 269], [270, 281], [282, 290], [291, 299], [300, 303], [304, 312], [313, 314], [314, 326], [327, 334], [335, 338], [339, 349], [349, 350], [351, 354], [354, 355]]}
{"doc_key": "ai-dev-91", "ner": [[3, 4, "person"], [10, 11, "conference"], [16, 20, "organisation"], [23, 30, "location"], [35, 35, "person"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[3, 4, 10, 11, "physical", "", false, false], [3, 4, 10, 11, "role", "", false, false], [3, 4, 16, 20, "role", "", false, false], [16, 20, 23, 30, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["En", "2000", ",", "Manuel", "Toharia", ",", "intervenant", "lors", "des", "pr\u00e9c\u00e9dentes", "Campus", "Parties", ",", "et", "directeur", "du", "mus\u00e9e", "des", "sciences", "Pr\u00edncipe", "Felipe", "de", "la", "Cit\u00e9", "des", "arts", "et", "des", "sciences", "de", "Valence", ",", "a", "sugg\u00e9r\u00e9", "\u00e0", "Ragageles", "d'", "\u00e9largir", "et", "de", "rendre", "plus", "international", "l'", "\u00e9v\u00e9nement", "en", "le", "d\u00e9pla\u00e7ant", "dans", "le", "c\u00e9l\u00e8bre", "mus\u00e9e", "."], "sentence-detokenized": "En 2000, Manuel Toharia, intervenant lors des pr\u00e9c\u00e9dentes Campus Parties, et directeur du mus\u00e9e des sciences Pr\u00edncipe Felipe de la Cit\u00e9 des arts et des sciences de Valence, a sugg\u00e9r\u00e9 \u00e0 Ragageles d'\u00e9largir et de rendre plus international l'\u00e9v\u00e9nement en le d\u00e9pla\u00e7ant dans le c\u00e9l\u00e8bre mus\u00e9e.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 15], [16, 23], [23, 24], [25, 36], [37, 41], [42, 45], [46, 57], [58, 64], [65, 72], [72, 73], [74, 76], [77, 86], [87, 89], [90, 95], [96, 99], [100, 108], [109, 117], [118, 124], [125, 127], [128, 130], [131, 135], [136, 139], [140, 144], [145, 147], [148, 151], [152, 160], [161, 163], [164, 171], [171, 172], [173, 174], [175, 182], [183, 184], [185, 194], [195, 197], [197, 204], [205, 207], [208, 210], [211, 217], [218, 222], [223, 236], [237, 239], [239, 248], [249, 251], [252, 254], [255, 264], [265, 269], [270, 272], [273, 280], [281, 286], [286, 287]]}
{"doc_key": "ai-dev-92", "ner": [[5, 8, "product"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["En", "20", "minutes", ",", "un", "syst\u00e8me", "de", "reconnaissance", "faciale", "identifie", "des", "informations", "personnelles", ",", "notamment", "le", "nom", "de", "famille", ",", "le", "num\u00e9ro", "d'", "identification", "et", "l'", "adresse", ",", "qui", "sont", "affich\u00e9es", "dans", "la", "rue", "sur", "un", "\u00e9cran", "publicitaire", "."], "sentence-detokenized": "En 20 minutes, un syst\u00e8me de reconnaissance faciale identifie des informations personnelles, notamment le nom de famille, le num\u00e9ro d'identification et l'adresse, qui sont affich\u00e9es dans la rue sur un \u00e9cran publicitaire.", "token2charspan": [[0, 2], [3, 5], [6, 13], [13, 14], [15, 17], [18, 25], [26, 28], [29, 43], [44, 51], [52, 61], [62, 65], [66, 78], [79, 91], [91, 92], [93, 102], [103, 105], [106, 109], [110, 112], [113, 120], [120, 121], [122, 124], [125, 131], [132, 134], [134, 148], [149, 151], [152, 154], [154, 161], [161, 162], [163, 166], [167, 171], [172, 181], [182, 186], [187, 189], [190, 193], [194, 197], [198, 200], [201, 206], [207, 219], [219, 220]]}
{"doc_key": "ai-dev-93", "ner": [[14, 16, "field"], [18, 18, "field"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Les", "recherches", "r\u00e9centes", "se", "sont", "de", "plus", "en", "plus", "concentr\u00e9es", "sur", "les", "algorithmes", "d'", "apprentissage", "non", "supervis\u00e9", "et", "semi-supervis\u00e9", "."], "sentence-detokenized": "Les recherches r\u00e9centes se sont de plus en plus concentr\u00e9es sur les algorithmes d'apprentissage non supervis\u00e9 et semi-supervis\u00e9.", "token2charspan": [[0, 3], [4, 14], [15, 23], [24, 26], [27, 31], [32, 34], [35, 39], [40, 42], [43, 47], [48, 59], [60, 63], [64, 67], [68, 79], [80, 82], [82, 95], [96, 99], [100, 109], [110, 112], [113, 127], [127, 128]]}
{"doc_key": "ai-dev-94", "ner": [[9, 9, "programlang"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Calcul", "de", "cet", "exemple", "\u00e0", "l'", "aide", "du", "code", "Python", ":"], "sentence-detokenized": "Calcul de cet exemple \u00e0 l'aide du code Python :", "token2charspan": [[0, 6], [7, 9], [10, 13], [14, 21], [22, 23], [24, 26], [26, 30], [31, 33], [34, 38], [39, 45], [46, 47]]}
{"doc_key": "ai-dev-95", "ner": [[9, 13, "task"], [18, 19, "field"], [21, 24, "algorithm"], [26, 26, "algorithm"], [30, 32, "algorithm"], [35, 36, "researcher"], [38, 39, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[21, 24, 18, 19, "part-of", "", false, false], [21, 24, 30, 32, "type-of", "", false, false], [21, 24, 35, 36, "origin", "", false, false], [21, 24, 38, 39, "origin", "", false, false], [26, 26, 21, 24, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Aujourd'hui", ",", "cependant", ",", "de", "nombreux", "aspects", "de", "la", "reconnaissance", "vocale", "ont", "\u00e9t\u00e9", "repris", "par", "une", "m\u00e9thode", "d'", "apprentissage", "profond", "appel\u00e9e", "m\u00e9moire", "\u00e0", "long", "terme", "(", "LSTM", ")", ",", "un", "r\u00e9seau", "neuronal", "r\u00e9current", "publi\u00e9", "par", "Sepp", "Hochreiter", "&", "J\u00fcrgen", "Schmidhuber", "en", "1997", "."], "sentence-detokenized": "Aujourd'hui, cependant, de nombreux aspects de la reconnaissance vocale ont \u00e9t\u00e9 repris par une m\u00e9thode d'apprentissage profond appel\u00e9e m\u00e9moire \u00e0 long terme (LSTM), un r\u00e9seau neuronal r\u00e9current publi\u00e9 par Sepp Hochreiter & J\u00fcrgen Schmidhuber en 1997.", "token2charspan": [[0, 11], [11, 12], [13, 22], [22, 23], [24, 26], [27, 35], [36, 43], [44, 46], [47, 49], [50, 64], [65, 71], [72, 75], [76, 79], [80, 86], [87, 90], [91, 94], [95, 102], [103, 105], [105, 118], [119, 126], [127, 134], [135, 142], [143, 144], [145, 149], [150, 155], [156, 157], [157, 161], [161, 162], [162, 163], [164, 166], [167, 173], [174, 182], [183, 192], [193, 199], [200, 203], [204, 208], [209, 219], [220, 221], [222, 228], [229, 240], [241, 243], [244, 248], [248, 249]]}
{"doc_key": "ai-dev-96", "ner": [[12, 12, "algorithm"], [20, 20, "algorithm"], [24, 24, "algorithm"], [32, 32, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[12, 12, 20, 20, "compare", "", false, false], [12, 12, 32, 32, "named", "same", false, false], [24, 24, 32, 32, "compare", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Dans", "les", "r\u00e9sultats", "exp\u00e9rimentaux", "pr\u00e9liminaires", "avec", "des", "ensembles", "de", "donn\u00e9es", "bruyants", ",", "BrownBoost", "a", "d\u00e9pass\u00e9", "l'", "erreur", "de", "g\u00e9n\u00e9ralisation", "d'", "AdaBoost", ";", "cependant", ",", "LogitBoost", "a", "donn\u00e9", "d'", "aussi", "bons", "r\u00e9sultats", "que", "BrownBoost", "."], "sentence-detokenized": "Dans les r\u00e9sultats exp\u00e9rimentaux pr\u00e9liminaires avec des ensembles de donn\u00e9es bruyants, BrownBoost a d\u00e9pass\u00e9 l'erreur de g\u00e9n\u00e9ralisation d'AdaBoost ; cependant, LogitBoost a donn\u00e9 d'aussi bons r\u00e9sultats que BrownBoost.", "token2charspan": [[0, 4], [5, 8], [9, 18], [19, 32], [33, 46], [47, 51], [52, 55], [56, 65], [66, 68], [69, 76], [77, 85], [85, 86], [87, 97], [98, 99], [100, 107], [108, 110], [110, 116], [117, 119], [120, 134], [135, 137], [137, 145], [146, 147], [148, 157], [157, 158], [159, 169], [170, 171], [172, 177], [178, 180], [180, 185], [186, 190], [191, 200], [201, 204], [205, 215], [215, 216]]}
{"doc_key": "ai-dev-97", "ner": [[0, 2, "algorithm"], [7, 9, "researcher"], [11, 11, "country"], [15, 18, "researcher"], [23, 24, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[0, 2, 7, 9, "part-of", "", false, false], [7, 9, 11, 11, "physical", "", false, false], [23, 24, 15, 18, "origin", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["La", "programmation", "\u00e9volutive", "a", "\u00e9t\u00e9", "introduite", "par", "Lawrence", "J.", "Fogel", "aux", "\u00c9tats-Unis", ",", "tandis", "que", "John", "Henry", "Holland", "a", "appel\u00e9", "sa", "m\u00e9thode", "un", "algorithme", "g\u00e9n\u00e9tique", "."], "sentence-detokenized": "La programmation \u00e9volutive a \u00e9t\u00e9 introduite par Lawrence J. Fogel aux \u00c9tats-Unis, tandis que John Henry Holland a appel\u00e9 sa m\u00e9thode un algorithme g\u00e9n\u00e9tique.", "token2charspan": [[0, 2], [3, 16], [17, 26], [27, 28], [29, 32], [33, 43], [44, 47], [48, 56], [57, 59], [60, 65], [66, 69], [70, 80], [80, 81], [82, 88], [89, 92], [93, 97], [98, 103], [104, 111], [112, 113], [114, 120], [121, 123], [124, 131], [132, 134], [135, 145], [146, 155], [155, 156]]}
{"doc_key": "ai-dev-98", "ner": [[7, 7, "researcher"], [9, 9, "researcher"], [15, 16, "researcher"], [18, 19, "researcher"], [21, 22, "researcher"], [24, 25, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[7, 7, 15, 16, "role", "", false, false], [7, 7, 18, 19, "role", "", false, false], [7, 7, 21, 22, "role", "", false, false], [7, 7, 24, 25, "role", "", false, false], [9, 9, 15, 16, "role", "", false, false], [9, 9, 18, 19, "role", "", false, false], [9, 9, 21, 22, "role", "", false, false], [9, 9, 24, 25, "role", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "sentence": ["Les", "calculs", "\u00e0", "l'", "envers", "effectu\u00e9s", "par", "Doug", ",", "Alan", "et", "leurs", "coll\u00e8gues", "(", "dont", "Marvin", "Minsky", ",", "Allen", "Newell", ",", "Edward", "Feigenbaum", "et", "John", "McCarthy", ")", "indiquaient", "que", "cet", "effort", "n\u00e9cessiterait", "entre", "1000", "et", "3000", "ann\u00e9es-personnes", ",", "bien", "au-del\u00e0", "du", "mod\u00e8le", "standard", "de", "projet", "universitaire", "."], "sentence-detokenized": "Les calculs \u00e0 l'envers effectu\u00e9s par Doug, Alan et leurs coll\u00e8gues (dont Marvin Minsky, Allen Newell, Edward Feigenbaum et John McCarthy) indiquaient que cet effort n\u00e9cessiterait entre 1000 et 3000 ann\u00e9es-personnes, bien au-del\u00e0 du mod\u00e8le standard de projet universitaire.", "token2charspan": [[0, 3], [4, 11], [12, 13], [14, 16], [16, 22], [23, 32], [33, 36], [37, 41], [41, 42], [43, 47], [48, 50], [51, 56], [57, 66], [67, 68], [68, 72], [73, 79], [80, 86], [86, 87], [88, 93], [94, 100], [100, 101], [102, 108], [109, 119], [120, 122], [123, 127], [128, 136], [136, 137], [138, 149], [150, 153], [154, 157], [158, 164], [165, 178], [179, 184], [185, 189], [190, 192], [193, 197], [198, 214], [214, 215], [216, 220], [221, 228], [229, 231], [232, 238], [239, 247], [248, 250], [251, 257], [258, 271], [271, 272]]}
{"doc_key": "ai-dev-99", "ner": [[8, 10, "metrics"], [15, 15, "metrics"], [18, 21, "metrics"], [26, 26, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[8, 10, 15, 15, "part-of", "implemented_in", false, false], [18, 21, 26, 26, "part-of", "implemented_in", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Les", "crit\u00e8res", "communs", "sont", "le", "crit\u00e8re", "de", "l'", "erreur", "quadratique", "moyenne", "mis", "en", "\u0153uvre", "dans", "MSECriterion", "et", "le", "crit\u00e8re", "d'", "entropie", "crois\u00e9e", "mis", "en", "\u0153uvre", "dans", "NLLCriterion", "."], "sentence-detokenized": "Les crit\u00e8res communs sont le crit\u00e8re de l'erreur quadratique moyenne mis en \u0153uvre dans MSECriterion et le crit\u00e8re d'entropie crois\u00e9e mis en \u0153uvre dans NLLCriterion.", "token2charspan": [[0, 3], [4, 12], [13, 20], [21, 25], [26, 28], [29, 36], [37, 39], [40, 42], [42, 48], [49, 60], [61, 68], [69, 72], [73, 75], [76, 81], [82, 86], [87, 99], [100, 102], [103, 105], [106, 113], [114, 116], [116, 124], [125, 132], [133, 136], [137, 139], [140, 145], [146, 150], [151, 163], [163, 164]]}
{"doc_key": "ai-dev-100", "ner": [[0, 1, "researcher"], [17, 17, "organisation"], [20, 32, "misc"], [41, 44, "conference"], [51, 51, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[0, 1, 17, 17, "role", "", false, false], [0, 1, 41, 44, "role", "", false, false], [0, 1, 51, 51, "role", "", false, false], [20, 32, 0, 1, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["M.", "Zurada", "a", "servi", "la", "profession", "d'", "ing\u00e9nieur", "en", "tant", "que", "b\u00e9n\u00e9vole", "de", "longue", "date", "de", "l'", "IEEE", ":", "en", "2014", ",", "il", "a", "\u00e9t\u00e9", "vice-pr\u00e9sident", "de", "l'", "IEEE", "charg\u00e9", "des", "activit\u00e9s", "techniques", "(", "TAB", "Chair", ")", ",", "pr\u00e9sident", "de", "l'", "IEEE", "Computational", "Intelligence", "Society", "en", "2004-05", "et", "membre", "de", "l'", "ADCOM", "en", "2009-14", ",", "2016-18", "et", "les", "ann\u00e9es", "pr\u00e9c\u00e9dentes", "."], "sentence-detokenized": "M. Zurada a servi la profession d'ing\u00e9nieur en tant que b\u00e9n\u00e9vole de longue date de l'IEEE : en 2014, il a \u00e9t\u00e9 vice-pr\u00e9sident de l'IEEE charg\u00e9 des activit\u00e9s techniques (TAB Chair), pr\u00e9sident de l'IEEE Computational Intelligence Society en 2004-05 et membre de l'ADCOM en 2009-14, 2016-18 et les ann\u00e9es pr\u00e9c\u00e9dentes.", "token2charspan": [[0, 2], [3, 9], [10, 11], [12, 17], [18, 20], [21, 31], [32, 34], [34, 43], [44, 46], [47, 51], [52, 55], [56, 64], [65, 67], [68, 74], [75, 79], [80, 82], [83, 85], [85, 89], [90, 91], [92, 94], [95, 99], [99, 100], [101, 103], [104, 105], [106, 109], [110, 124], [125, 127], [128, 130], [130, 134], [135, 141], [142, 145], [146, 155], [156, 166], [167, 168], [168, 171], [172, 177], [177, 178], [178, 179], [180, 189], [190, 192], [193, 195], [195, 199], [200, 213], [214, 226], [227, 234], [235, 237], [238, 245], [246, 248], [249, 255], [256, 258], [259, 261], [261, 266], [267, 269], [270, 277], [277, 278], [279, 286], [287, 289], [290, 293], [294, 300], [301, 312], [312, 313]]}
{"doc_key": "ai-dev-101", "ner": [[4, 9, "field"], [15, 15, "field"], [20, 21, "field"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[15, 15, 4, 9, "part-of", "", false, false], [20, 21, 4, 9, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["En", "g\u00e9n\u00e9ral", ",", "la", "linguistique", "computationnelle", "fait", "appel", "\u00e0", "la", "participation", "de", "linguistes", ",", "d'", "informaticiens", ",", "d'", "experts", "en", "intelligence", "artificielle", ",", "de", "math\u00e9maticiens", ",", "de", "logiciens", ",", "de", "philosophes", ",", "de", "sp\u00e9cialistes", "des", "sciences", "cognitives", ",", "de", "psychologues", "cognitifs", ",", "de", "psycholinguistes", ",", "d'", "anthropologues", "et", "de", "neuroscientifiques", ",", "entre", "autres", "."], "sentence-detokenized": "En g\u00e9n\u00e9ral, la linguistique computationnelle fait appel \u00e0 la participation de linguistes, d'informaticiens, d'experts en intelligence artificielle, de math\u00e9maticiens, de logiciens, de philosophes, de sp\u00e9cialistes des sciences cognitives, de psychologues cognitifs, de psycholinguistes, d'anthropologues et de neuroscientifiques, entre autres.", "token2charspan": [[0, 2], [3, 10], [10, 11], [12, 14], [15, 27], [28, 44], [45, 49], [50, 55], [56, 57], [58, 60], [61, 74], [75, 77], [78, 88], [88, 89], [90, 92], [92, 106], [106, 107], [108, 110], [110, 117], [118, 120], [121, 133], [134, 146], [146, 147], [148, 150], [151, 165], [165, 166], [167, 169], [170, 179], [179, 180], [181, 183], [184, 195], [195, 196], [197, 199], [200, 212], [213, 216], [217, 225], [226, 236], [236, 237], [238, 240], [241, 253], [254, 263], [263, 264], [265, 267], [268, 284], [284, 285], [286, 288], [288, 302], [303, 305], [306, 308], [309, 327], [327, 328], [329, 334], [335, 341], [341, 342]]}
{"doc_key": "ai-dev-102", "ner": [[5, 8, "algorithm"], [11, 13, "algorithm"], [16, 19, "algorithm"]], "ner_mapping_to_source": [0, 1, 2], "relations": [], "relations_mapping_to_source": [], "sentence": ["Des", "techniques", "telles", "que", "les", "r\u00e9seaux", "de", "Markov", "dynamiques", ",", "les", "r\u00e9seaux", "neuronaux", "convolutionnels", "et", "la", "m\u00e9moire", "\u00e0", "long", "terme", "sont", "souvent", "utilis\u00e9es", "pour", "exploiter", "les", "corr\u00e9lations", "entre", "les", "images", "."], "sentence-detokenized": "Des techniques telles que les r\u00e9seaux de Markov dynamiques, les r\u00e9seaux neuronaux convolutionnels et la m\u00e9moire \u00e0 long terme sont souvent utilis\u00e9es pour exploiter les corr\u00e9lations entre les images.", "token2charspan": [[0, 3], [4, 14], [15, 21], [22, 25], [26, 29], [30, 37], [38, 40], [41, 47], [48, 58], [58, 59], [60, 63], [64, 71], [72, 81], [82, 97], [98, 100], [101, 103], [104, 111], [112, 113], [114, 118], [119, 124], [125, 129], [130, 137], [138, 147], [148, 152], [153, 162], [163, 166], [167, 179], [180, 185], [186, 189], [190, 196], [196, 197]]}
{"doc_key": "ai-dev-103", "ner": [[0, 0, "product"], [4, 5, "product"]], "ner_mapping_to_source": [0, 1], "relations": [[0, 0, 4, 5, "type-of", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Unimate", "\u00e9tait", "le", "premier", "robot", "industriel", ","], "sentence-detokenized": "Unimate \u00e9tait le premier robot industriel,", "token2charspan": [[0, 7], [8, 13], [14, 16], [17, 24], [25, 30], [31, 41], [41, 42]]}
{"doc_key": "ai-dev-104", "ner": [[1, 2, "researcher"], [4, 5, "researcher"], [7, 8, "researcher"], [11, 13, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[1, 2, 11, 13, "win-defeat", "", false, false], [4, 5, 11, 13, "win-defeat", "", false, false], [7, 8, 11, 13, "win-defeat", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Avec", "Geoffrey", "Hinton", "et", "Yann", "LeCun", ",", "Bengio", "a", "remport\u00e9", "le", "prix", "Turing", "2018", "."], "sentence-detokenized": "Avec Geoffrey Hinton et Yann LeCun, Bengio a remport\u00e9 le prix Turing 2018.", "token2charspan": [[0, 4], [5, 13], [14, 20], [21, 23], [24, 28], [29, 34], [34, 35], [36, 42], [43, 44], [45, 53], [54, 56], [57, 61], [62, 68], [69, 73], [73, 74]]}
{"doc_key": "ai-dev-105", "ner": [[9, 9, "country"], [22, 25, "misc"], [32, 33, "organisation"], [37, 38, "person"], [40, 41, "person"], [51, 53, "misc"], [58, 58, "country"], [65, 65, "country"]], "ner_mapping_to_source": [0, 1, 3, 4, 5, 6, 7, 8], "relations": [[22, 25, 9, 9, "physical", "filmed_in", false, false], [37, 38, 32, 33, "role", "host", false, false], [40, 41, 32, 33, "role", "reporter", false, false], [51, 53, 9, 9, "physical", "filmed_in", false, false], [51, 53, 58, 58, "physical", "distributed_in", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["D'", "autres", "s\u00e9ries", "ont", "\u00e9t\u00e9", "tourn\u00e9es", "sur", "le", "site", "britannique", "pour", "des", "secteurs", "sp\u00e9cifiques", "du", "march\u00e9", "mondial", ",", "notamment", "deux", "s\u00e9ries", "de", "Robot", "Wars", "Extreme", "Warriors", "avec", "des", "concurrents", "am\u00e9ricains", "pour", "le", "r\u00e9seau", "TNN", "(", "anim\u00e9es", "par", "Mick", "Foley", "avec", "Rebecca", "Grant", "comme", "reporter", "de", "stand", ")", ",", "deux", "s\u00e9ries", "de", "Dutch", "Robot", "Wars", "pour", "la", "distribution", "aux", "Pays-Bas", "et", "une", "s\u00e9rie", "unique", "pour", "l'", "Allemagne", "."], "sentence-detokenized": "D'autres s\u00e9ries ont \u00e9t\u00e9 tourn\u00e9es sur le site britannique pour des secteurs sp\u00e9cifiques du march\u00e9 mondial, notamment deux s\u00e9ries de Robot Wars Extreme Warriors avec des concurrents am\u00e9ricains pour le r\u00e9seau TNN (anim\u00e9es par Mick Foley avec Rebecca Grant comme reporter de stand), deux s\u00e9ries de Dutch Robot Wars pour la distribution aux Pays-Bas et une s\u00e9rie unique pour l'Allemagne.", "token2charspan": [[0, 2], [2, 8], [9, 15], [16, 19], [20, 23], [24, 32], [33, 36], [37, 39], [40, 44], [45, 56], [57, 61], [62, 65], [66, 74], [75, 86], [87, 89], [90, 96], [97, 104], [104, 105], [106, 115], [116, 120], [121, 127], [128, 130], [131, 136], [137, 141], [142, 149], [150, 158], [159, 163], [164, 167], [168, 179], [180, 190], [191, 195], [196, 198], [199, 205], [206, 209], [210, 211], [211, 218], [219, 222], [223, 227], [228, 233], [234, 238], [239, 246], [247, 252], [253, 258], [259, 267], [268, 270], [271, 276], [276, 277], [277, 278], [279, 283], [284, 290], [291, 293], [294, 299], [300, 305], [306, 310], [311, 315], [316, 318], [319, 331], [332, 335], [336, 344], [345, 347], [348, 351], [352, 357], [358, 364], [365, 369], [370, 372], [372, 381], [381, 382]]}
{"doc_key": "ai-dev-106", "ner": [[9, 10, "researcher"], [15, 15, "product"], [32, 34, "product"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[9, 10, 15, 15, "role", "", false, false], [32, 34, 15, 15, "usage", "", true, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Pendant", "de", "nombreuses", "ann\u00e9es", "\u00e0", "partir", "de", "1986", ",", "Miller", "a", "dirig\u00e9", "le", "d\u00e9veloppement", "de", "WordNet", ",", "une", "grande", "r\u00e9f\u00e9rence", "\u00e9lectronique", "lisible", "par", "ordinateur", "et", "utilisable", "dans", "des", "applications", "telles", "que", "les", "moteurs", "de", "recherche", "."], "sentence-detokenized": "Pendant de nombreuses ann\u00e9es \u00e0 partir de 1986, Miller a dirig\u00e9 le d\u00e9veloppement de WordNet, une grande r\u00e9f\u00e9rence \u00e9lectronique lisible par ordinateur et utilisable dans des applications telles que les moteurs de recherche.", "token2charspan": [[0, 7], [8, 10], [11, 21], [22, 28], [29, 30], [31, 37], [38, 40], [41, 45], [45, 46], [47, 53], [54, 55], [56, 62], [63, 65], [66, 79], [80, 82], [83, 90], [90, 91], [92, 95], [96, 102], [103, 112], [113, 125], [126, 133], [134, 137], [138, 148], [149, 151], [152, 162], [163, 167], [168, 171], [172, 184], [185, 191], [192, 195], [196, 199], [200, 207], [208, 210], [211, 220], [220, 221]]}
{"doc_key": "ai-dev-107", "ner": [[4, 6, "algorithm"], [9, 12, "algorithm"], [20, 21, "researcher"], [23, 26, "organisation"], [30, 33, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[4, 6, 20, 21, "origin", "", false, false], [4, 6, 30, 33, "win-defeat", "", false, false], [9, 12, 20, 21, "origin", "", false, false], [9, 12, 30, 33, "win-defeat", "", false, false], [20, 21, 23, 26, "physical", "", false, false], [20, 21, 23, 26, "role", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["Depuis", "2009", ",", "les", "r\u00e9seaux", "neuronaux", "r\u00e9currents", "et", "les", "r\u00e9seaux", "neuronaux", "feedforward", "profonds", "d\u00e9velopp\u00e9s", "dans", "le", "groupe", "de", "recherche", "de", "J\u00fcrgen", "Schmidhuber", "au", "Swiss", "AI", "Lab", "IDSIA", "ont", "remport\u00e9", "plusieurs", "concours", "internationaux", "d'", "\u00e9criture", "..."], "sentence-detokenized": "Depuis 2009, les r\u00e9seaux neuronaux r\u00e9currents et les r\u00e9seaux neuronaux feedforward profonds d\u00e9velopp\u00e9s dans le groupe de recherche de J\u00fcrgen Schmidhuber au Swiss AI Lab IDSIA ont remport\u00e9 plusieurs concours internationaux d'\u00e9criture...", "token2charspan": [[0, 6], [7, 11], [11, 12], [13, 16], [17, 24], [25, 34], [35, 45], [46, 48], [49, 52], [53, 60], [61, 70], [71, 82], [83, 91], [92, 102], [103, 107], [108, 110], [111, 117], [118, 120], [121, 130], [131, 133], [134, 140], [141, 152], [153, 155], [156, 161], [162, 164], [165, 168], [169, 174], [175, 178], [179, 187], [188, 197], [198, 206], [207, 221], [222, 224], [224, 232], [232, 235]]}
{"doc_key": "ai-dev-108", "ner": [[5, 7, "programlang"], [13, 13, "programlang"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Le", "logiciel", "est", "impl\u00e9ment\u00e9", "en", "C", "+", "+", "et", "il", "est", "envelopp\u00e9", "pour", "Python", "."], "sentence-detokenized": "Le logiciel est impl\u00e9ment\u00e9 en C + + et il est envelopp\u00e9 pour Python.", "token2charspan": [[0, 2], [3, 11], [12, 15], [16, 26], [27, 29], [30, 31], [32, 33], [34, 35], [36, 38], [39, 41], [42, 45], [46, 55], [56, 60], [61, 67], [67, 68]]}
{"doc_key": "ai-dev-109", "ner": [[7, 8, "country"], [13, 15, "misc"], [21, 22, "misc"], [39, 40, "misc"], [42, 42, "misc"], [45, 45, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[21, 22, 7, 8, "temporal", "", false, false], [21, 22, 13, 15, "artifact", "", false, false], [21, 22, 45, 45, "physical", "", false, false], [42, 42, 39, 40, "named", "", false, false], [42, 42, 45, 45, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["En", "1857", ",", "\u00e0", "la", "demande", "du", "shogunat", "Tokugawa", ",", "un", "groupe", "d'", "ing\u00e9nieurs", "n\u00e9erlandais", "a", "commenc\u00e9", "\u00e0", "travailler", "sur", "le", "Nagasaki", "Yotetsusho", ",", "une", "fonderie", "et", "un", "chantier", "naval", "modernes", ",", "de", "style", "occidental", ",", "pr\u00e8s", "de", "la", "colonie", "n\u00e9erlandaise", "de", "Dejima", ",", "\u00e0", "Nagasaki", "."], "sentence-detokenized": "En 1857, \u00e0 la demande du shogunat Tokugawa, un groupe d'ing\u00e9nieurs n\u00e9erlandais a commenc\u00e9 \u00e0 travailler sur le Nagasaki Yotetsusho, une fonderie et un chantier naval modernes, de style occidental, pr\u00e8s de la colonie n\u00e9erlandaise de Dejima, \u00e0 Nagasaki.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 10], [11, 13], [14, 21], [22, 24], [25, 33], [34, 42], [42, 43], [44, 46], [47, 53], [54, 56], [56, 66], [67, 78], [79, 80], [81, 89], [90, 91], [92, 102], [103, 106], [107, 109], [110, 118], [119, 129], [129, 130], [131, 134], [135, 143], [144, 146], [147, 149], [150, 158], [159, 164], [165, 173], [173, 174], [175, 177], [178, 183], [184, 194], [194, 195], [196, 200], [201, 203], [204, 206], [207, 214], [215, 227], [228, 230], [231, 237], [237, 238], [239, 240], [241, 249], [249, 250]]}
{"doc_key": "ai-dev-110", "ner": [[7, 9, "metrics"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Nous", "faisons", "au", "mieux", "en", "mesurant", "l'", "erreur", "quadratique", "moyenne", "entre", "mathy", "/", "math", "et", "math", "\\", "hat", "{f}", "(", "x", ";", "D", ")", "/", "math", ":", "nous", "voulons", "que", "math", "(", "y", "-", "\\", "hat", "{f}", "(", "x", ";", "D", ")", ")", "^", "2", "/", "maths", "soit", "minimale", ",", "aussi", "bien", "pour", "mathx", "_", "1", ",", "\\", "points", ",", "x", "_n", "/", "maths", "que", "pour", "les", "points", "hors", "de", "notre", "\u00e9chantillon", "."], "sentence-detokenized": "Nous faisons au mieux en mesurant l'erreur quadratique moyenne entre mathy / math et math\\ hat {f} (x ; D) / math : nous voulons que math (y -\\ hat {f} (x ; D)) ^ 2 / maths soit minimale, aussi bien pour mathx _ 1,\\ points, x _n / maths que pour les points hors de notre \u00e9chantillon.", "token2charspan": [[0, 4], [5, 12], [13, 15], [16, 21], [22, 24], [25, 33], [34, 36], [36, 42], [43, 54], [55, 62], [63, 68], [69, 74], [75, 76], [77, 81], [82, 84], [85, 89], [89, 90], [91, 94], [95, 98], [99, 100], [100, 101], [102, 103], [104, 105], [105, 106], [107, 108], [109, 113], [114, 115], [116, 120], [121, 128], [129, 132], [133, 137], [138, 139], [139, 140], [141, 142], [142, 143], [144, 147], [148, 151], [152, 153], [153, 154], [155, 156], [157, 158], [158, 159], [159, 160], [161, 162], [163, 164], [165, 166], [167, 172], [173, 177], [178, 186], [186, 187], [188, 193], [194, 198], [199, 203], [204, 209], [210, 211], [212, 213], [213, 214], [214, 215], [216, 222], [222, 223], [224, 225], [226, 228], [229, 230], [231, 236], [237, 240], [241, 245], [246, 249], [250, 256], [257, 261], [262, 264], [265, 270], [271, 282], [282, 283]]}
{"doc_key": "ai-dev-111", "ner": [[4, 5, "researcher"], [14, 16, "organisation"], [25, 29, "product"], [42, 43, "task"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[4, 5, 14, 16, "role", "", false, false], [25, 29, 14, 16, "temporal", "", false, false], [25, 29, 42, 43, "related-to", "performs", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Il", "a", "ensuite", "invit\u00e9", "M.", "Wydner", "\u00e0", "assister", "\u00e0", "la", "r\u00e9union", "annuelle", "de", "l'", "American", "Translators", "Association", "au", "mois", "d'", "octobre", "suivant", ",", "o\u00f9", "le", "syst\u00e8me", "de", "traduction", "automatique", "Weidner", "a", "\u00e9t\u00e9", "salu\u00e9", "comme", "une", "perc\u00e9e", "esp\u00e9r\u00e9e", "dans", "le", "domaine", "de", "la", "traduction", "automatique", "."], "sentence-detokenized": "Il a ensuite invit\u00e9 M. Wydner \u00e0 assister \u00e0 la r\u00e9union annuelle de l'American Translators Association au mois d'octobre suivant, o\u00f9 le syst\u00e8me de traduction automatique Weidner a \u00e9t\u00e9 salu\u00e9 comme une perc\u00e9e esp\u00e9r\u00e9e dans le domaine de la traduction automatique.", "token2charspan": [[0, 2], [3, 4], [5, 12], [13, 19], [20, 22], [23, 29], [30, 31], [32, 40], [41, 42], [43, 45], [46, 53], [54, 62], [63, 65], [66, 68], [68, 76], [77, 88], [89, 100], [101, 103], [104, 108], [109, 111], [111, 118], [119, 126], [126, 127], [128, 130], [131, 133], [134, 141], [142, 144], [145, 155], [156, 167], [168, 175], [176, 177], [178, 181], [182, 187], [188, 193], [194, 197], [198, 204], [205, 212], [213, 217], [218, 220], [221, 228], [229, 231], [232, 234], [235, 245], [246, 257], [257, 258]]}
{"doc_key": "ai-dev-112", "ner": [[3, 13, "conference"], [15, 15, "conference"], [21, 21, "organisation"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[15, 15, 3, 13, "named", "", false, false], [15, 15, 3, 13, "temporal", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Lors", "de", "la", "conf\u00e9rence", "2018", "sur", "les", "syst\u00e8mes", "de", "traitement", "de", "l'", "information", "neuronale", "(", "NeurIPS", ")", ",", "des", "chercheurs", "de", "Google", "ont", "pr\u00e9sent\u00e9", "ces", "travaux", "."], "sentence-detokenized": "Lors de la conf\u00e9rence 2018 sur les syst\u00e8mes de traitement de l'information neuronale (NeurIPS), des chercheurs de Google ont pr\u00e9sent\u00e9 ces travaux.", "token2charspan": [[0, 4], [5, 7], [8, 10], [11, 21], [22, 26], [27, 30], [31, 34], [35, 43], [44, 46], [47, 57], [58, 60], [61, 63], [63, 74], [75, 84], [85, 86], [86, 93], [93, 94], [94, 95], [96, 99], [100, 110], [111, 113], [114, 120], [121, 124], [125, 133], [134, 137], [138, 145], [145, 146]]}
{"doc_key": "ai-dev-113", "ner": [[1, 3, "algorithm"], [6, 7, "algorithm"], [13, 17, "metrics"], [22, 25, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[1, 3, 6, 7, "usage", "", false, false], [6, 7, 13, 17, "related-to", "", true, false], [13, 17, 22, 25, "related-to", "", true, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["L'", "algorithme", "de", "Baum-Welch", "utilise", "l'", "algorithme", "EM", "bien", "connu", "pour", "trouver", "l'", "estimation", "du", "maximum", "de", "vraisemblance", "des", "param\u00e8tres", "d'", "un", "mod\u00e8le", "de", "Markov", "cach\u00e9", ",", "\u00e9tant", "donn\u00e9", "un", "ensemble", "de", "vecteurs", "caract\u00e9ristiques", "observ\u00e9s", "."], "sentence-detokenized": "L'algorithme de Baum-Welch utilise l'algorithme EM bien connu pour trouver l'estimation du maximum de vraisemblance des param\u00e8tres d'un mod\u00e8le de Markov cach\u00e9, \u00e9tant donn\u00e9 un ensemble de vecteurs caract\u00e9ristiques observ\u00e9s.", "token2charspan": [[0, 2], [2, 12], [13, 15], [16, 26], [27, 34], [35, 37], [37, 47], [48, 50], [51, 55], [56, 61], [62, 66], [67, 74], [75, 77], [77, 87], [88, 90], [91, 98], [99, 101], [102, 115], [116, 119], [120, 130], [131, 133], [133, 135], [136, 142], [143, 145], [146, 152], [153, 158], [158, 159], [160, 165], [166, 171], [172, 174], [175, 183], [184, 186], [187, 195], [196, 212], [213, 221], [221, 222]]}
{"doc_key": "ai-dev-114", "ner": [[8, 8, "product"], [10, 10, "product"], [32, 34, "misc"], [40, 51, "product"], [58, 58, "programlang"], [55, 66, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[8, 8, 10, 10, "compare", "", false, false], [32, 34, 10, 10, "part-of", "", false, false], [40, 51, 10, 10, "part-of", "", false, false], [55, 66, 10, 10, "part-of", "", false, false], [55, 66, 58, 58, "general-affiliation", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": [")", "En", "plus", "des", "informations", "taxonomiques", "contenues", "dans", "OpenCyc", ",", "ResearchCyc", "comprend", "beaucoup", "plus", "de", "connaissances", "s\u00e9mantiques", "(", "c'est-\u00e0-dire", "des", "faits", "et", "des", "r\u00e8gles", "empiriques", "suppl\u00e9mentaires", ")", "impliquant", "les", "concepts", "de", "sa", "base", "de", "connaissances", ";", "il", "comprend", "\u00e9galement", "un", "grand", "lexique", ",", "des", "outils", "d'", "analyse", "et", "de", "g\u00e9n\u00e9ration", "en", "anglais", ",", "et", "des", "interfaces", "bas\u00e9es", "sur", "Java", "pour", "l'", "\u00e9dition", "et", "l'", "interrogation", "des", "connaissances", "."], "sentence-detokenized": ") En plus des informations taxonomiques contenues dans OpenCyc, ResearchCyc comprend beaucoup plus de connaissances s\u00e9mantiques (c'est-\u00e0-dire des faits et des r\u00e8gles empiriques suppl\u00e9mentaires) impliquant les concepts de sa base de connaissances ; il comprend \u00e9galement un grand lexique, des outils d'analyse et de g\u00e9n\u00e9ration en anglais, et des interfaces bas\u00e9es sur Java pour l'\u00e9dition et l'interrogation des connaissances.", "token2charspan": [[0, 1], [2, 4], [5, 9], [10, 13], [14, 26], [27, 39], [40, 49], [50, 54], [55, 62], [62, 63], [64, 75], [76, 84], [85, 93], [94, 98], [99, 101], [102, 115], [116, 127], [128, 129], [129, 141], [142, 145], [146, 151], [152, 154], [155, 158], [159, 165], [166, 176], [177, 192], [192, 193], [194, 204], [205, 208], [209, 217], [218, 220], [221, 223], [224, 228], [229, 231], [232, 245], [246, 247], [248, 250], [251, 259], [260, 269], [270, 272], [273, 278], [279, 286], [286, 287], [288, 291], [292, 298], [299, 301], [301, 308], [309, 311], [312, 314], [315, 325], [326, 328], [329, 336], [336, 337], [338, 340], [341, 344], [345, 355], [356, 362], [363, 366], [367, 371], [372, 376], [377, 379], [379, 386], [387, 389], [390, 392], [392, 405], [406, 409], [410, 423], [423, 424]]}
{"doc_key": "ai-dev-115", "ner": [[0, 3, "algorithm"], [8, 10, "task"], [13, 15, "field"], [18, 20, "field"], [23, 26, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[0, 3, 8, 10, "type-of", "", false, false], [8, 10, 13, 15, "part-of", "task_part_of_field", false, false], [8, 10, 18, 20, "part-of", "task_part_of_field", false, false], [8, 10, 23, 26, "part-of", "task_part_of_field", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["La", "transform\u00e9e", "de", "Hough", "est", "une", "technique", "d'", "extraction", "de", "caract\u00e9ristiques", "utilis\u00e9e", "en", "analyse", "d'", "images", ",", "en", "vision", "par", "ordinateur", "et", "en", "traitement", "num\u00e9rique", "des", "images", "."], "sentence-detokenized": "La transform\u00e9e de Hough est une technique d'extraction de caract\u00e9ristiques utilis\u00e9e en analyse d'images, en vision par ordinateur et en traitement num\u00e9rique des images.", "token2charspan": [[0, 2], [3, 14], [15, 17], [18, 23], [24, 27], [28, 31], [32, 41], [42, 44], [44, 54], [55, 57], [58, 74], [75, 83], [84, 86], [87, 94], [95, 97], [97, 103], [103, 104], [105, 107], [108, 114], [115, 118], [119, 129], [130, 132], [133, 135], [136, 146], [147, 156], [157, 160], [161, 167], [167, 168]]}
{"doc_key": "ai-dev-116", "ner": [[5, 5, "product"], [7, 11, "product"], [17, 17, "organisation"], [21, 21, "product"], [23, 24, "researcher"], [31, 32, "organisation"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[5, 5, 7, 11, "named", "", false, false], [5, 5, 17, 17, "artifact", "", false, false], [5, 5, 21, 21, "origin", "developed_from", false, false], [21, 21, 23, 24, "artifact", "", false, false], [31, 32, 17, 17, "role", "supported", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["En", "1978", ",", "le", "robot", "PUMA", "(", "Programmable", "Universal", "Machine", "for", "Assembly", ")", "a", "\u00e9t\u00e9", "d\u00e9velopp\u00e9", "par", "Unimation", "\u00e0", "partir", "de", "Vicarm", "(", "Victor", "Scheinman", ")", "et", "avec", "le", "soutien", "de", "General", "Motors", "."], "sentence-detokenized": "En 1978, le robot PUMA (Programmable Universal Machine for Assembly) a \u00e9t\u00e9 d\u00e9velopp\u00e9 par Unimation \u00e0 partir de Vicarm (Victor Scheinman) et avec le soutien de General Motors.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 11], [12, 17], [18, 22], [23, 24], [24, 36], [37, 46], [47, 54], [55, 58], [59, 67], [67, 68], [69, 70], [71, 74], [75, 84], [85, 88], [89, 98], [99, 100], [101, 107], [108, 110], [111, 117], [118, 119], [119, 125], [126, 135], [135, 136], [137, 139], [140, 144], [145, 147], [148, 155], [156, 158], [159, 166], [167, 173], [173, 174]]}
{"doc_key": "ai-dev-117", "ner": [[1, 1, "algorithm"], [8, 9, "researcher"], [11, 12, "researcher"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[1, 1, 8, 9, "origin", "", false, false], [1, 1, 11, 12, "origin", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Le", "LSTM", "a", "\u00e9t\u00e9", "propos\u00e9", "en", "1997", "par", "Sepp", "Hochreiter", "et", "J\u00fcrgen", "Schmidhuber", "."], "sentence-detokenized": "Le LSTM a \u00e9t\u00e9 propos\u00e9 en 1997 par Sepp Hochreiter et J\u00fcrgen Schmidhuber.", "token2charspan": [[0, 2], [3, 7], [8, 9], [10, 13], [14, 21], [22, 24], [25, 29], [30, 33], [34, 38], [39, 49], [50, 52], [53, 59], [60, 71], [71, 72]]}
{"doc_key": "ai-dev-118", "ner": [[8, 10, "metrics"], [16, 18, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Les", "quatre", "r\u00e9sultats", "peuvent", "\u00eatre", "formul\u00e9s", "dans", "un", "tableau", "de", "contingence", "2", "\u00d7", "2", "ou", "une", "matrice", "de", "confusion", ",", "comme", "suit", ":"], "sentence-detokenized": "Les quatre r\u00e9sultats peuvent \u00eatre formul\u00e9s dans un tableau de contingence 2 \u00d7 2 ou une matrice de confusion, comme suit :", "token2charspan": [[0, 3], [4, 10], [11, 20], [21, 28], [29, 33], [34, 42], [43, 47], [48, 50], [51, 58], [59, 61], [62, 73], [74, 75], [76, 77], [78, 79], [80, 82], [83, 86], [87, 94], [95, 97], [98, 107], [107, 108], [109, 114], [115, 119], [120, 121]]}
{"doc_key": "ai-dev-119", "ner": [[10, 10, "conference"], [14, 15, "conference"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "a", "\u00e9galement", "beaucoup", "contribu\u00e9", "\u00e0", "la", "cr\u00e9ation", "de", "l'", "ELRA", "et", "de", "la", "conf\u00e9rence", "LREC", "."], "sentence-detokenized": "Il a \u00e9galement beaucoup contribu\u00e9 \u00e0 la cr\u00e9ation de l'ELRA et de la conf\u00e9rence LREC.", "token2charspan": [[0, 2], [3, 4], [5, 14], [15, 23], [24, 33], [34, 35], [36, 38], [39, 47], [48, 50], [51, 53], [53, 57], [58, 60], [61, 63], [64, 66], [67, 77], [78, 82], [82, 83]]}
{"doc_key": "ai-dev-120", "ner": [[14, 16, "misc"], [20, 21, "product"]], "ner_mapping_to_source": [0, 1], "relations": [[20, 21, 14, 16, "general-affiliation", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Une", "application", "populaire", "des", "robots", "en", "s\u00e9rie", "dans", "l'", "industrie", "d'", "aujourd'hui", "est", "le", "robot", "d'", "assemblage", "pick-and-place", ",", "appel\u00e9", "robot", "SCARA", ",", "qui", "a", "quatre", "degr\u00e9s", "de", "libert\u00e9", "."], "sentence-detokenized": "Une application populaire des robots en s\u00e9rie dans l'industrie d'aujourd'hui est le robot d'assemblage pick-and-place, appel\u00e9 robot SCARA, qui a quatre degr\u00e9s de libert\u00e9.", "token2charspan": [[0, 3], [4, 15], [16, 25], [26, 29], [30, 36], [37, 39], [40, 45], [46, 50], [51, 53], [53, 62], [63, 65], [65, 76], [77, 80], [81, 83], [84, 89], [90, 92], [92, 102], [103, 117], [117, 118], [119, 125], [126, 131], [132, 137], [137, 138], [139, 142], [143, 144], [145, 151], [152, 158], [159, 161], [162, 169], [169, 170]]}
{"doc_key": "ai-dev-121", "ner": [[16, 22, "conference"], [24, 24, "conference"], [28, 31, "conference"], [39, 39, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[24, 24, 16, 22, "named", "", false, false], [39, 39, 28, 31, "named", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Il", "a", "\u00e9t\u00e9", "l'", "un", "des", "membres", "fondateurs", "et", "l'", "ancien", "pr\u00e9sident", "(", "2006-2008", ")", "du", "Special", "Interest", "Group", "on", "Web", "as", "Corpus", "(", "SIGWAC", ")", "de", "l'", "Association", "for", "Computational", "Linguistics", "et", "l'", "un", "des", "organisateurs", "fondateurs", "de", "SENSEVAL", "."], "sentence-detokenized": "Il a \u00e9t\u00e9 l'un des membres fondateurs et l'ancien pr\u00e9sident (2006-2008) du Special Interest Group on Web as Corpus (SIGWAC) de l'Association for Computational Linguistics et l'un des organisateurs fondateurs de SENSEVAL.", "token2charspan": [[0, 2], [3, 4], [5, 8], [9, 11], [11, 13], [14, 17], [18, 25], [26, 36], [37, 39], [40, 42], [42, 48], [49, 58], [59, 60], [60, 69], [69, 70], [71, 73], [74, 81], [82, 90], [91, 96], [97, 99], [100, 103], [104, 106], [107, 113], [114, 115], [115, 121], [121, 122], [123, 125], [126, 128], [128, 139], [140, 143], [144, 157], [158, 169], [170, 172], [173, 175], [175, 177], [178, 181], [182, 195], [196, 206], [207, 209], [210, 218], [218, 219]]}
{"doc_key": "ai-dev-122", "ner": [[5, 5, "product"], [8, 9, "product"]], "ner_mapping_to_source": [0, 1], "relations": [[8, 9, 5, 5, "part-of", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["En", "tant", "que", "plateforme", ",", "LinguaStream", "fournit", "une", "API", "Java", "\u00e9tendue", "."], "sentence-detokenized": "En tant que plateforme, LinguaStream fournit une API Java \u00e9tendue.", "token2charspan": [[0, 2], [3, 7], [8, 11], [12, 22], [22, 23], [24, 36], [37, 44], [45, 48], [49, 52], [53, 57], [58, 65], [65, 66]]}
{"doc_key": "ai-dev-123", "ner": [[17, 17, "programlang"], [21, 24, "misc"], [29, 32, "product"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[17, 17, 29, 32, "type-of", "", false, false], [21, 24, 29, 32, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Le", "kit", "de", "robot", "est", "bas\u00e9", "sur", "Android", ",", "et", "il", "est", "programm\u00e9", "\u00e0", "l'", "aide", "de", "Java", ",", "de", "l'", "interface", "de", "programmation", "Blocks", ",", "ou", "d'", "autres", "syst\u00e8mes", "de", "programmation", "Android", "."], "sentence-detokenized": "Le kit de robot est bas\u00e9 sur Android, et il est programm\u00e9 \u00e0 l'aide de Java, de l'interface de programmation Blocks, ou d'autres syst\u00e8mes de programmation Android.", "token2charspan": [[0, 2], [3, 6], [7, 9], [10, 15], [16, 19], [20, 24], [25, 28], [29, 36], [36, 37], [38, 40], [41, 43], [44, 47], [48, 57], [58, 59], [60, 62], [62, 66], [67, 69], [70, 74], [74, 75], [76, 78], [79, 81], [81, 90], [91, 93], [94, 107], [108, 114], [114, 115], [116, 118], [119, 121], [121, 127], [128, 136], [137, 139], [140, 153], [154, 161], [161, 162]]}
{"doc_key": "ai-dev-124", "ner": [[13, 15, "algorithm"], [19, 21, "algorithm"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["La", "m\u00e9thode", "de", "d\u00e9finition", "de", "la", "liste", "li\u00e9e", "sp\u00e9cifie", "l'", "utilisation", "d'", "une", "recherche", "en", "profondeur", "ou", "d'", "une", "recherche", "en", "largeur", "."], "sentence-detokenized": "La m\u00e9thode de d\u00e9finition de la liste li\u00e9e sp\u00e9cifie l'utilisation d'une recherche en profondeur ou d'une recherche en largeur.", "token2charspan": [[0, 2], [3, 10], [11, 13], [14, 24], [25, 27], [28, 30], [31, 36], [37, 41], [42, 50], [51, 53], [53, 64], [65, 67], [67, 70], [71, 80], [81, 83], [84, 94], [95, 97], [98, 100], [100, 103], [104, 113], [114, 116], [117, 124], [124, 125]]}
{"doc_key": "ai-dev-125", "ner": [[25, 27, "task"], [32, 35, "task"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Ces", "r\u00e9gions", "pourraient", "signaler", "la", "pr\u00e9sence", "d'", "objets", "ou", "de", "parties", "d'", "objets", "dans", "le", "domaine", "de", "l'", "image", ",", "avec", "une", "application", "\u00e0", "la", "reconnaissance", "d'", "objets", "et", "/", "ou", "au", "suivi", "vid\u00e9o", "d'", "objets", "."], "sentence-detokenized": "Ces r\u00e9gions pourraient signaler la pr\u00e9sence d'objets ou de parties d'objets dans le domaine de l'image, avec une application \u00e0 la reconnaissance d'objets et/ou au suivi vid\u00e9o d'objets.", "token2charspan": [[0, 3], [4, 11], [12, 22], [23, 31], [32, 34], [35, 43], [44, 46], [46, 52], [53, 55], [56, 58], [59, 66], [67, 69], [69, 75], [76, 80], [81, 83], [84, 91], [92, 94], [95, 97], [97, 102], [102, 103], [104, 108], [109, 112], [113, 124], [125, 126], [127, 129], [130, 144], [145, 147], [147, 153], [154, 156], [156, 157], [157, 159], [160, 162], [163, 168], [169, 174], [175, 177], [177, 183], [183, 184]]}
{"doc_key": "ai-dev-126", "ner": [[3, 4, "algorithm"], [6, 6, "product"], [15, 15, "misc"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[6, 6, 3, 4, "type-of", "", false, false], [6, 6, 15, 15, "topic", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Un", "exemple", "de", "r\u00e9seau", "s\u00e9mantique", "est", "WordNet", ",", "une", "base", "de", "donn\u00e9es", "lexicale", "de", "l'", "anglais", "."], "sentence-detokenized": "Un exemple de r\u00e9seau s\u00e9mantique est WordNet, une base de donn\u00e9es lexicale de l'anglais.", "token2charspan": [[0, 2], [3, 10], [11, 13], [14, 20], [21, 31], [32, 35], [36, 43], [43, 44], [45, 48], [49, 53], [54, 56], [57, 64], [65, 73], [74, 76], [77, 79], [79, 86], [86, 87]]}
{"doc_key": "ai-dev-127", "ner": [[0, 2, "task"], [9, 9, "field"], [13, 14, "field"], [24, 31, "task"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[0, 2, 9, 9, "part-of", "", false, false], [0, 2, 13, 14, "named", "same", false, false], [0, 2, 13, 14, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["La", "reconnaissance", "vocale", "est", "un", "sous-domaine", "interdisciplinaire", "de", "l'", "informatique", "et", "de", "la", "linguistique", "computationnelle", "qui", "d\u00e9veloppe", "des", "m\u00e9thodologies", "et", "des", "technologies", "permettant", "la", "reconnaissance", "et", "la", "traduction", "de", "la", "langue", "parl\u00e9e", "en", "texte", "par", "des", "ordinateurs", "."], "sentence-detokenized": "La reconnaissance vocale est un sous-domaine interdisciplinaire de l'informatique et de la linguistique computationnelle qui d\u00e9veloppe des m\u00e9thodologies et des technologies permettant la reconnaissance et la traduction de la langue parl\u00e9e en texte par des ordinateurs.", "token2charspan": [[0, 2], [3, 17], [18, 24], [25, 28], [29, 31], [32, 44], [45, 63], [64, 66], [67, 69], [69, 81], [82, 84], [85, 87], [88, 90], [91, 103], [104, 120], [121, 124], [125, 134], [135, 138], [139, 152], [153, 155], [156, 159], [160, 172], [173, 183], [184, 186], [187, 201], [202, 204], [205, 207], [208, 218], [219, 221], [222, 224], [225, 231], [232, 238], [239, 241], [242, 247], [248, 251], [252, 255], [256, 267], [267, 268]]}
{"doc_key": "ai-dev-128", "ner": [[1, 2, "field"], [14, 15, "misc"], [21, 24, "field"], [27, 27, "task"], [30, 32, "task"], [57, 57, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[1, 2, 57, 57, "named", "same", false, false], [21, 24, 1, 2, "part-of", "subfield", false, false], [27, 27, 1, 2, "part-of", "", false, false], [27, 27, 21, 24, "part-of", "", false, false], [30, 32, 21, 24, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["L'", "intelligence", "artificielle", "a", "retenu", "la", "plus", "grande", "attention", "en", "ce", "qui", "concerne", "l'", "ontologie", "appliqu\u00e9e", "dans", "des", "sous-domaines", "comme", "le", "traitement", "du", "langage", "naturel", "dans", "la", "machine", "et", "la", "repr\u00e9sentation", "des", "connaissances", ",", "mais", "les", "\u00e9diteurs", "d'", "ontologie", "sont", "souvent", "utilis\u00e9s", "dans", "une", "s\u00e9rie", "de", "domaines", "comme", "l'", "\u00e9ducation", "sans", "l'", "intention", "de", "contribuer", "\u00e0", "l'", "IA", "."], "sentence-detokenized": "L'intelligence artificielle a retenu la plus grande attention en ce qui concerne l'ontologie appliqu\u00e9e dans des sous-domaines comme le traitement du langage naturel dans la machine et la repr\u00e9sentation des connaissances, mais les \u00e9diteurs d'ontologie sont souvent utilis\u00e9s dans une s\u00e9rie de domaines comme l'\u00e9ducation sans l'intention de contribuer \u00e0 l'IA.", "token2charspan": [[0, 2], [2, 14], [15, 27], [28, 29], [30, 36], [37, 39], [40, 44], [45, 51], [52, 61], [62, 64], [65, 67], [68, 71], [72, 80], [81, 83], [83, 92], [93, 102], [103, 107], [108, 111], [112, 125], [126, 131], [132, 134], [135, 145], [146, 148], [149, 156], [157, 164], [165, 169], [170, 172], [173, 180], [181, 183], [184, 186], [187, 201], [202, 205], [206, 219], [219, 220], [221, 225], [226, 229], [230, 238], [239, 241], [241, 250], [251, 255], [256, 263], [264, 272], [273, 277], [278, 281], [282, 287], [288, 290], [291, 299], [300, 305], [306, 308], [308, 317], [318, 322], [323, 325], [325, 334], [335, 337], [338, 348], [349, 350], [351, 353], [353, 355], [355, 356]]}
{"doc_key": "ai-dev-129", "ner": [[14, 17, "algorithm"], [20, 21, "algorithm"]], "ner_mapping_to_source": [0, 1], "relations": [[14, 17, 20, 21, "related-to", "", true, false]], "relations_mapping_to_source": [0], "sentence": ["Cette", "r\u00e8gle", "de", "mise", "\u00e0", "jour", "est", "en", "fait", "la", "mise", "\u00e0", "jour", "par", "descente", "de", "gradient", "stochastique", "pour", "la", "r\u00e9gression", "lin\u00e9aire", "."], "sentence-detokenized": "Cette r\u00e8gle de mise \u00e0 jour est en fait la mise \u00e0 jour par descente de gradient stochastique pour la r\u00e9gression lin\u00e9aire.", "token2charspan": [[0, 5], [6, 11], [12, 14], [15, 19], [20, 21], [22, 26], [27, 30], [31, 33], [34, 38], [39, 41], [42, 46], [47, 48], [49, 53], [54, 57], [58, 66], [67, 69], [70, 78], [79, 91], [92, 96], [97, 99], [100, 110], [111, 119], [119, 120]]}
{"doc_key": "ai-dev-130", "ner": [[6, 12, "organisation"], [16, 19, "organisation"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "a", "\u00e9t\u00e9", "\u00e9lu", "\u00e0", "l'", "Acad\u00e9mie", "am\u00e9ricaine", "des", "arts", "et", "des", "sciences", "et", "\u00e0", "l'", "Acad\u00e9mie", "nationale", "des", "sciences", "et", "a", "re\u00e7u", "une", "s\u00e9rie", "de", "prix", ":"], "sentence-detokenized": "Il a \u00e9t\u00e9 \u00e9lu \u00e0 l'Acad\u00e9mie am\u00e9ricaine des arts et des sciences et \u00e0 l'Acad\u00e9mie nationale des sciences et a re\u00e7u une s\u00e9rie de prix :", "token2charspan": [[0, 2], [3, 4], [5, 8], [9, 12], [13, 14], [15, 17], [17, 25], [26, 36], [37, 40], [41, 45], [46, 48], [49, 52], [53, 61], [62, 64], [65, 66], [67, 69], [69, 77], [78, 87], [88, 91], [92, 100], [101, 103], [104, 105], [106, 110], [111, 114], [115, 120], [121, 123], [124, 128], [129, 130]]}
{"doc_key": "ai-dev-131", "ner": [[11, 11, "organisation"], [16, 17, "person"], [19, 21, "person"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[11, 11, 16, 17, "related-to", "written_about_by", false, false], [11, 11, 19, 21, "related-to", "written_about_by", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["L'", "\u00e9cole", "de", "pens\u00e9e", "la", "plus", "r\u00e9cente", "sur", "la", "strat\u00e9gie", "de", "Honda", "a", "\u00e9t\u00e9", "propos\u00e9e", "par", "Gary", "Hamel", "et", "C.", "K.", "Prahalad", "en", "1989", "."], "sentence-detokenized": "L'\u00e9cole de pens\u00e9e la plus r\u00e9cente sur la strat\u00e9gie de Honda a \u00e9t\u00e9 propos\u00e9e par Gary Hamel et C. K. Prahalad en 1989.", "token2charspan": [[0, 2], [2, 7], [8, 10], [11, 17], [18, 20], [21, 25], [26, 33], [34, 37], [38, 40], [41, 50], [51, 53], [54, 59], [60, 61], [62, 65], [66, 74], [75, 78], [79, 83], [84, 89], [90, 92], [93, 95], [96, 98], [99, 107], [108, 110], [111, 115], [115, 116]]}
{"doc_key": "ai-dev-132", "ner": [[2, 2, "metrics"], [6, 7, "metrics"], [19, 19, "metrics"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[2, 2, 6, 7, "related-to", "calculates", true, false], [2, 2, 19, 19, "compare", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Alors", "que", "BLEU", "calcule", "simplement", "la", "pr\u00e9cision", "d'", "un", "gramme", "en", "ajoutant", "un", "poids", "\u00e9gal", "\u00e0", "chacun", ",", "le", "NIST", "calcule", "\u00e9galement", "le", "degr\u00e9", "d'", "information", "d'", "un", "gramme", "particulier", "."], "sentence-detokenized": "Alors que BLEU calcule simplement la pr\u00e9cision d'un gramme en ajoutant un poids \u00e9gal \u00e0 chacun, le NIST calcule \u00e9galement le degr\u00e9 d'information d'un gramme particulier.", "token2charspan": [[0, 5], [6, 9], [10, 14], [15, 22], [23, 33], [34, 36], [37, 46], [47, 49], [49, 51], [52, 58], [59, 61], [62, 70], [71, 73], [74, 79], [80, 84], [85, 86], [87, 93], [93, 94], [95, 97], [98, 102], [103, 110], [111, 120], [121, 123], [124, 129], [130, 132], [132, 143], [144, 146], [146, 148], [149, 155], [156, 167], [167, 168]]}
{"doc_key": "ai-dev-133", "ner": [[6, 22, "misc"], [10, 13, "conference"], [15, 18, "conference"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[6, 22, 10, 13, "temporal", "", false, false], [15, 18, 10, 13, "named", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Il", "a", "\u00e9t\u00e9", "honor\u00e9", "par", "le", "prix", "2019", "de", "l'", "Association", "for", "Computational", "Linguistics", "(", "ACL", ")", "pour", "l'", "ensemble", "de", "ses", "r\u00e9alisations", "."], "sentence-detokenized": "Il a \u00e9t\u00e9 honor\u00e9 par le prix 2019 de l'Association for Computational Linguistics (ACL) pour l'ensemble de ses r\u00e9alisations.", "token2charspan": [[0, 2], [3, 4], [5, 8], [9, 15], [16, 19], [20, 22], [23, 27], [28, 32], [33, 35], [36, 38], [38, 49], [50, 53], [54, 67], [68, 79], [80, 81], [81, 84], [84, 85], [86, 90], [91, 93], [93, 101], [102, 104], [105, 108], [109, 121], [121, 122]]}
{"doc_key": "ai-dev-134", "ner": [[0, 0, "researcher"], [5, 10, "organisation"], [12, 12, "organisation"], [17, 21, "conference"], [23, 23, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[0, 0, 5, 10, "role", "", false, false], [0, 0, 17, 21, "role", "", false, false], [12, 12, 5, 10, "named", "", false, false], [23, 23, 17, 21, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Sycara", "est", "membre", "de", "l'", "Institute", "of", "Electrical", "and", "Electronics", "Engineers", "(", "IEEE", ")", "et", "de", "l'", "American", "Association", "for", "Artificial", "Intelligence", "(", "AAAI", ")", "."], "sentence-detokenized": "Sycara est membre de l'Institute of Electrical and Electronics Engineers (IEEE) et de l'American Association for Artificial Intelligence (AAAI).", "token2charspan": [[0, 6], [7, 10], [11, 17], [18, 20], [21, 23], [23, 32], [33, 35], [36, 46], [47, 50], [51, 62], [63, 72], [73, 74], [74, 78], [78, 79], [80, 82], [83, 85], [86, 88], [88, 96], [97, 108], [109, 112], [113, 123], [124, 136], [137, 138], [138, 142], [142, 143], [143, 144]]}
{"doc_key": "ai-dev-135", "ner": [[2, 2, "product"], [11, 15, "misc"]], "ner_mapping_to_source": [0, 1], "relations": [[2, 2, 11, 15, "related-to", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Le", "code", "MATLAB", "suivant", "d\u00e9montre", "une", "solution", "concr\u00e8te", "pour", "r\u00e9soudre", "le", "syst\u00e8me", "d'", "\u00e9quations", "non", "lin\u00e9aires", "pr\u00e9sent\u00e9", "dans", "la", "section", "pr\u00e9c\u00e9dente", ":", "Voir", "aussi"], "sentence-detokenized": "Le code MATLAB suivant d\u00e9montre une solution concr\u00e8te pour r\u00e9soudre le syst\u00e8me d'\u00e9quations non lin\u00e9aires pr\u00e9sent\u00e9 dans la section pr\u00e9c\u00e9dente : Voir aussi", "token2charspan": [[0, 2], [3, 7], [8, 14], [15, 22], [23, 31], [32, 35], [36, 44], [45, 53], [54, 58], [59, 67], [68, 70], [71, 78], [79, 81], [81, 90], [91, 94], [95, 104], [105, 113], [114, 118], [119, 121], [122, 129], [130, 140], [141, 142], [143, 147], [148, 153]]}
{"doc_key": "ai-dev-136", "ner": [[6, 10, "product"], [21, 22, "field"], [47, 49, "field"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[6, 10, 21, 22, "related-to", "trained_by", true, false], [6, 10, 47, 49, "related-to", "trained_by", true, false], [21, 22, 47, 49, "compare", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Dans", "de", "nombreux", "cas", ",", "les", "syst\u00e8mes", "de", "reconnaissance", "des", "formes", "sont", "form\u00e9s", "\u00e0", "partir", "de", "donn\u00e9es", "d'", "apprentissage", "\u00e9tiquet\u00e9es", "(", "apprentissage", "supervis\u00e9", ")", ",", "mais", "lorsqu'", "aucune", "donn\u00e9e", "\u00e9tiquet\u00e9e", "n'", "est", "disponible", ",", "d'", "autres", "algorithmes", "peuvent", "\u00eatre", "utilis\u00e9s", "pour", "d\u00e9couvrir", "des", "formes", "pr\u00e9c\u00e9demment", "inconnues", "(", "apprentissage", "non", "supervis\u00e9", ")", "."], "sentence-detokenized": "Dans de nombreux cas, les syst\u00e8mes de reconnaissance des formes sont form\u00e9s \u00e0 partir de donn\u00e9es d'apprentissage \u00e9tiquet\u00e9es (apprentissage supervis\u00e9), mais lorsqu'aucune donn\u00e9e \u00e9tiquet\u00e9e n'est disponible, d'autres algorithmes peuvent \u00eatre utilis\u00e9s pour d\u00e9couvrir des formes pr\u00e9c\u00e9demment inconnues (apprentissage non supervis\u00e9).", "token2charspan": [[0, 4], [5, 7], [8, 16], [17, 20], [20, 21], [22, 25], [26, 34], [35, 37], [38, 52], [53, 56], [57, 63], [64, 68], [69, 75], [76, 77], [78, 84], [85, 87], [88, 95], [96, 98], [98, 111], [112, 122], [123, 124], [124, 137], [138, 147], [147, 148], [148, 149], [150, 154], [155, 162], [162, 168], [169, 175], [176, 185], [186, 188], [188, 191], [192, 202], [202, 203], [204, 206], [206, 212], [213, 224], [225, 232], [233, 237], [238, 246], [247, 251], [252, 261], [262, 265], [266, 272], [273, 285], [286, 295], [296, 297], [297, 310], [311, 314], [315, 324], [324, 325], [325, 326]]}
{"doc_key": "ai-dev-137", "ner": [[9, 11, "researcher"], [13, 13, "country"], [30, 31, "field"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[9, 11, 13, 13, "physical", "", false, false], [9, 11, 30, 31, "related-to", "works_with", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Il", "a", "\u00e9t\u00e9", "utilis\u00e9", "pour", "la", "premi\u00e8re", "fois", "par", "Lawrence", "J.", "Fogel", "aux", "\u00c9tats-Unis", "en", "1960", "afin", "d'", "utiliser", "l'", "\u00e9volution", "simul\u00e9e", "comme", "processus", "d'", "apprentissage", "visant", "\u00e0", "g\u00e9n\u00e9rer", "une", "intelligence", "artificielle", "."], "sentence-detokenized": "Il a \u00e9t\u00e9 utilis\u00e9 pour la premi\u00e8re fois par Lawrence J. Fogel aux \u00c9tats-Unis en 1960 afin d'utiliser l'\u00e9volution simul\u00e9e comme processus d'apprentissage visant \u00e0 g\u00e9n\u00e9rer une intelligence artificielle.", "token2charspan": [[0, 2], [3, 4], [5, 8], [9, 16], [17, 21], [22, 24], [25, 33], [34, 38], [39, 42], [43, 51], [52, 54], [55, 60], [61, 64], [65, 75], [76, 78], [79, 83], [84, 88], [89, 91], [91, 99], [100, 102], [102, 111], [112, 119], [120, 125], [126, 135], [136, 138], [138, 151], [152, 158], [159, 160], [161, 168], [169, 172], [173, 185], [186, 198], [198, 199]]}
{"doc_key": "ai-dev-138", "ner": [[0, 3, "field"], [13, 14, "field"], [18, 19, "field"], [22, 24, "field"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[0, 3, 13, 14, "part-of", "", false, false], [18, 19, 13, 14, "part-of", "", false, false], [22, 24, 13, 14, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["L'", "apprentissage", "par", "renforcement", "est", "l'", "un", "des", "trois", "paradigmes", "fondamentaux", "de", "l'", "apprentissage", "automatique", ",", "avec", "l'", "apprentissage", "supervis\u00e9", "et", "l'", "apprentissage", "non", "supervis\u00e9", "."], "sentence-detokenized": "L'apprentissage par renforcement est l'un des trois paradigmes fondamentaux de l'apprentissage automatique, avec l'apprentissage supervis\u00e9 et l'apprentissage non supervis\u00e9.", "token2charspan": [[0, 2], [2, 15], [16, 19], [20, 32], [33, 36], [37, 39], [39, 41], [42, 45], [46, 51], [52, 62], [63, 75], [76, 78], [79, 81], [81, 94], [95, 106], [106, 107], [108, 112], [113, 115], [115, 128], [129, 138], [139, 141], [142, 144], [144, 157], [158, 161], [162, 171], [171, 172]]}
{"doc_key": "ai-dev-139", "ner": [[5, 6, "field"], [14, 14, "programlang"], [38, 39, "field"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[5, 6, 38, 39, "usage", "applies", false, false], [14, 14, 38, 39, "usage", "applies", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Dans", "ce", "cas", ",", "le", "cloud", "computing", "et", "le", "langage", "de", "programmation", "open", "source", "R", "peuvent", "aider", "les", "petites", "banques", "\u00e0", "adopter", "l'", "analyse", "des", "risques", "et", "\u00e0", "soutenir", "la", "surveillance", "au", "niveau", "des", "agences", "en", "appliquant", "l'", "analyse", "pr\u00e9dictive", "."], "sentence-detokenized": "Dans ce cas, le cloud computing et le langage de programmation open source R peuvent aider les petites banques \u00e0 adopter l'analyse des risques et \u00e0 soutenir la surveillance au niveau des agences en appliquant l'analyse pr\u00e9dictive.", "token2charspan": [[0, 4], [5, 7], [8, 11], [11, 12], [13, 15], [16, 21], [22, 31], [32, 34], [35, 37], [38, 45], [46, 48], [49, 62], [63, 67], [68, 74], [75, 76], [77, 84], [85, 90], [91, 94], [95, 102], [103, 110], [111, 112], [113, 120], [121, 123], [123, 130], [131, 134], [135, 142], [143, 145], [146, 147], [148, 156], [157, 159], [160, 172], [173, 175], [176, 182], [183, 186], [187, 194], [195, 197], [198, 208], [209, 211], [211, 218], [219, 229], [229, 230]]}
{"doc_key": "ai-dev-140", "ner": [[10, 11, "researcher"], [20, 22, "algorithm"], [24, 25, "researcher"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[10, 11, 24, 25, "named", "same", false, false], [20, 22, 10, 11, "origin", "proves_function", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Une", "des", "premi\u00e8res", "versions", "du", "th\u00e9or\u00e8me", "a", "\u00e9t\u00e9", "prouv\u00e9e", "par", "George", "Cybenko", "en", "1989", "pour", "les", "fonctions", "d'", "activation", "de", "la", "fonction", "sigmo\u00efde", ".", "Cybenko", "G.", "(", "1989", ")", ",", "2", "(", "4", ")", ",", "303-314", "."], "sentence-detokenized": "Une des premi\u00e8res versions du th\u00e9or\u00e8me a \u00e9t\u00e9 prouv\u00e9e par George Cybenko en 1989 pour les fonctions d'activation de la fonction sigmo\u00efde. Cybenko G. (1989), 2 (4), 303-314.", "token2charspan": [[0, 3], [4, 7], [8, 17], [18, 26], [27, 29], [30, 38], [39, 40], [41, 44], [45, 52], [53, 56], [57, 63], [64, 71], [72, 74], [75, 79], [80, 84], [85, 88], [89, 98], [99, 101], [101, 111], [112, 114], [115, 117], [118, 126], [127, 135], [135, 136], [137, 144], [145, 147], [148, 149], [149, 153], [153, 154], [154, 155], [156, 157], [158, 159], [159, 160], [160, 161], [161, 162], [163, 170], [170, 171]]}
{"doc_key": "ai-dev-141", "ner": [[9, 10, "algorithm"], [13, 13, "metrics"], [17, 21, "metrics"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[13, 13, 9, 10, "part-of", "", false, false], [17, 21, 13, 13, "named", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Dans", "ce", "processus", ",", "connu", "sous", "le", "nom", "de", "validation", "crois\u00e9e", ",", "l'", "EQM", "est", "souvent", "appel\u00e9e", "erreur", "quadratique", "moyenne", "de", "pr\u00e9diction", ",", "et", "est", "calcul\u00e9e", "comme", "suit"], "sentence-detokenized": "Dans ce processus, connu sous le nom de validation crois\u00e9e, l'EQM est souvent appel\u00e9e erreur quadratique moyenne de pr\u00e9diction, et est calcul\u00e9e comme suit", "token2charspan": [[0, 4], [5, 7], [8, 17], [17, 18], [19, 24], [25, 29], [30, 32], [33, 36], [37, 39], [40, 50], [51, 58], [58, 59], [60, 62], [62, 65], [66, 69], [70, 77], [78, 85], [86, 92], [93, 104], [105, 112], [113, 115], [116, 126], [126, 127], [128, 130], [131, 134], [135, 143], [144, 149], [150, 154]]}
{"doc_key": "ai-dev-142", "ner": [[1, 1, "task"], [7, 10, "task"], [12, 12, "task"], [26, 28, "field"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[1, 1, 7, 10, "compare", "", false, false], [7, 10, 26, 28, "part-of", "", false, false], [12, 12, 7, 10, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["L'", "OMR", "se", "distingue", "g\u00e9n\u00e9ralement", "de", "la", "reconnaissance", "optique", "de", "caract\u00e8res", "(", "OCR", ")", "par", "le", "fait", "qu'", "elle", "ne", "n\u00e9cessite", "pas", "de", "moteur", "complexe", "de", "reconnaissance", "des", "formes", "."], "sentence-detokenized": "L'OMR se distingue g\u00e9n\u00e9ralement de la reconnaissance optique de caract\u00e8res (OCR) par le fait qu'elle ne n\u00e9cessite pas de moteur complexe de reconnaissance des formes.", "token2charspan": [[0, 2], [2, 5], [6, 8], [9, 18], [19, 31], [32, 34], [35, 37], [38, 52], [53, 60], [61, 63], [64, 74], [75, 76], [76, 79], [79, 80], [81, 84], [85, 87], [88, 92], [93, 96], [96, 100], [101, 103], [104, 113], [114, 117], [118, 120], [121, 127], [128, 136], [137, 139], [140, 154], [155, 158], [159, 165], [165, 166]]}
{"doc_key": "ai-dev-143", "ner": [[11, 11, "location"], [14, 14, "location"], [18, 18, "location"], [21, 22, "location"], [25, 26, "location"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[14, 14, 18, 18, "physical", "", false, false], [21, 22, 14, 14, "physical", "", false, false], [25, 26, 14, 14, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["En", "2018", "et", "2019", ",", "le", "championnat", "s'", "est", "d\u00e9roul\u00e9", "\u00e0", "Houston", "et", "\u00e0", "D\u00e9troit", ",", "dans", "le", "Michigan", ",", "au", "TCF", "Center", "et", "au", "Ford", "Field", "."], "sentence-detokenized": "En 2018 et 2019, le championnat s'est d\u00e9roul\u00e9 \u00e0 Houston et \u00e0 D\u00e9troit, dans le Michigan, au TCF Center et au Ford Field.", "token2charspan": [[0, 2], [3, 7], [8, 10], [11, 15], [15, 16], [17, 19], [20, 31], [32, 34], [34, 37], [38, 45], [46, 47], [48, 55], [56, 58], [59, 60], [61, 68], [68, 69], [70, 74], [75, 77], [78, 86], [86, 87], [88, 90], [91, 94], [95, 101], [102, 104], [105, 107], [108, 112], [113, 118], [118, 119]]}
{"doc_key": "ai-dev-144", "ner": [[0, 1, "task"], [11, 12, "task"], [15, 16, "task"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[11, 12, 0, 1, "part-of", "", false, false], [15, 16, 0, 1, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["La", "classification", "peut", "\u00eatre", "consid\u00e9r\u00e9e", "comme", "deux", "probl\u00e8mes", "distincts", ":", "la", "classification", "binaire", "et", "la", "classification", "multi-classes", "."], "sentence-detokenized": "La classification peut \u00eatre consid\u00e9r\u00e9e comme deux probl\u00e8mes distincts : la classification binaire et la classification multi-classes.", "token2charspan": [[0, 2], [3, 17], [18, 22], [23, 27], [28, 38], [39, 44], [45, 49], [50, 59], [60, 69], [70, 71], [72, 74], [75, 89], [90, 97], [98, 100], [101, 103], [104, 118], [119, 132], [132, 133]]}
{"doc_key": "ai-dev-145", "ner": [[3, 4, "product"], [8, 9, "product"], [12, 13, "product"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[8, 9, 3, 4, "type-of", "", false, false], [12, 13, 3, 4, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Deux", "exemples", "de", "robots", "parall\u00e8les", "populaires", "sont", "la", "plate-forme", "Stewart", "et", "le", "robot", "Delta", "."], "sentence-detokenized": "Deux exemples de robots parall\u00e8les populaires sont la plate-forme Stewart et le robot Delta.", "token2charspan": [[0, 4], [5, 13], [14, 16], [17, 23], [24, 34], [35, 45], [46, 50], [51, 53], [54, 65], [66, 73], [74, 76], [77, 79], [80, 85], [86, 91], [91, 92]]}
{"doc_key": "ai-dev-146", "ner": [[4, 7, "algorithm"], [23, 23, "algorithm"]], "ner_mapping_to_source": [0, 1], "relations": [[4, 7, 23, 23, "part-of", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["(", "N\u00e9anmoins", ",", "la", "fonction", "d'", "activation", "ReLU", ",", "qui", "est", "indiff\u00e9renciable", "\u00e0", "0", ",", "est", "devenue", "assez", "populaire", ",", "par", "exemple", "dans", "AlexNet", ")", "."], "sentence-detokenized": "(N\u00e9anmoins, la fonction d'activation ReLU, qui est indiff\u00e9renciable \u00e0 0, est devenue assez populaire, par exemple dans AlexNet).", "token2charspan": [[0, 1], [1, 10], [10, 11], [12, 14], [15, 23], [24, 26], [26, 36], [37, 41], [41, 42], [43, 46], [47, 50], [51, 67], [68, 69], [70, 71], [71, 72], [73, 76], [77, 84], [85, 90], [91, 100], [100, 101], [102, 105], [106, 113], [114, 118], [119, 126], [126, 127], [127, 128]]}
{"doc_key": "ai-dev-147", "ner": [[1, 2, "metrics"], [11, 13, "task"], [20, 20, "task"], [24, 26, "task"], [30, 32, "task"], [35, 36, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[1, 2, 35, 36, "named", "", true, false], [11, 13, 1, 2, "usage", "", true, false], [20, 20, 11, 13, "part-of", "", false, false], [24, 26, 11, 13, "part-of", "", false, false], [30, 32, 11, 13, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Le", "score", "F", "est", "souvent", "utilis\u00e9", "dans", "le", "domaine", "de", "la", "recherche", "d'", "informations", "pour", "mesurer", "la", "performance", "de", "la", "recherche", ",", "de", "la", "classification", "des", "documents", "et", "de", "la", "classification", "des", "requ\u00eates", ".", "Le", "F_beta", "est", "donc", "largement", "utilis\u00e9", "."], "sentence-detokenized": "Le score F est souvent utilis\u00e9 dans le domaine de la recherche d'informations pour mesurer la performance de la recherche, de la classification des documents et de la classification des requ\u00eates. Le F_beta est donc largement utilis\u00e9.", "token2charspan": [[0, 2], [3, 8], [9, 10], [11, 14], [15, 22], [23, 30], [31, 35], [36, 38], [39, 46], [47, 49], [50, 52], [53, 62], [63, 65], [65, 77], [78, 82], [83, 90], [91, 93], [94, 105], [106, 108], [109, 111], [112, 121], [121, 122], [123, 125], [126, 128], [129, 143], [144, 147], [148, 157], [158, 160], [161, 163], [164, 166], [167, 181], [182, 185], [186, 194], [194, 195], [196, 198], [199, 205], [206, 209], [210, 214], [215, 224], [225, 232], [232, 233]]}
{"doc_key": "ai-dev-148", "ner": [[21, 23, "algorithm"], [25, 25, "algorithm"], [29, 30, "algorithm"], [32, 32, "algorithm"], [36, 38, "algorithm"], [40, 40, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[25, 25, 21, 23, "named", "", false, false], [32, 32, 29, 30, "named", "", false, false], [40, 40, 36, 38, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Pour", "ce", "faire", ",", "on", "mod\u00e9lise", "le", "signal", "re\u00e7u", ",", "puis", "on", "utilise", "une", "m\u00e9thode", "d'", "estimation", "statistique", "telle", "que", "le", "maximum", "de", "vraisemblance", "(", "ML", ")", ",", "le", "vote", "majoritaire", "(", "MV", ")", "ou", "le", "maximum", "a", "posteriori", "(", "MAP", ")", "pour", "d\u00e9cider", "quelle", "cible", "de", "la", "biblioth\u00e8que", "correspond", "le", "mieux", "au", "mod\u00e8le", "construit", "\u00e0", "l'", "aide", "du", "signal", "re\u00e7u", "."], "sentence-detokenized": "Pour ce faire, on mod\u00e9lise le signal re\u00e7u, puis on utilise une m\u00e9thode d'estimation statistique telle que le maximum de vraisemblance (ML), le vote majoritaire (MV) ou le maximum a posteriori (MAP) pour d\u00e9cider quelle cible de la biblioth\u00e8que correspond le mieux au mod\u00e8le construit \u00e0 l'aide du signal re\u00e7u.", "token2charspan": [[0, 4], [5, 7], [8, 13], [13, 14], [15, 17], [18, 26], [27, 29], [30, 36], [37, 41], [41, 42], [43, 47], [48, 50], [51, 58], [59, 62], [63, 70], [71, 73], [73, 83], [84, 95], [96, 101], [102, 105], [106, 108], [109, 116], [117, 119], [120, 133], [134, 135], [135, 137], [137, 138], [138, 139], [140, 142], [143, 147], [148, 159], [160, 161], [161, 163], [163, 164], [165, 167], [168, 170], [171, 178], [179, 180], [181, 191], [192, 193], [193, 196], [196, 197], [198, 202], [203, 210], [211, 217], [218, 223], [224, 226], [227, 229], [230, 242], [243, 253], [254, 256], [257, 262], [263, 265], [266, 272], [273, 282], [283, 284], [285, 287], [287, 291], [292, 294], [295, 301], [302, 306], [306, 307]]}
{"doc_key": "ai-dev-149", "ner": [[0, 2, "researcher"], [5, 5, "misc"], [7, 7, "field"], [9, 12, "university"], [17, 17, "misc"], [19, 22, "field"], [23, 25, "university"], [30, 30, "misc"], [32, 32, "field"], [35, 37, "university"], [44, 53, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10], "relations": [[0, 2, 9, 12, "physical", "", false, false], [0, 2, 9, 12, "role", "", false, false], [0, 2, 23, 25, "physical", "", false, false], [0, 2, 23, 25, "role", "", false, false], [0, 2, 35, 37, "physical", "", false, false], [0, 2, 35, 37, "role", "", false, false], [5, 5, 0, 2, "origin", "", false, false], [5, 5, 7, 7, "topic", "", false, false], [17, 17, 0, 2, "origin", "", false, false], [17, 17, 19, 22, "topic", "", false, false], [30, 30, 0, 2, "origin", "", false, false], [30, 30, 32, 32, "topic", "", false, false], [44, 53, 30, 30, "named", "", true, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12], "sentence": ["M.", "Sowa", "a", "obtenu", "un", "BS", "en", "math\u00e9matiques", "au", "Massachusetts", "Institute", "of", "Technology", "en", "1962", ",", "un", "MA", "en", "sciences", "appliqu\u00e9es", "\u00e0", "l'", "Universit\u00e9", "de", "Harvard", "en", "1966", "et", "un", "doctorat", "en", "informatique", "\u00e0", "la", "Vrije", "Universiteit", "Brussel", "en", "1999", "sur", "une", "th\u00e8se", "intitul\u00e9e", "Knowledge", "Representation", ":", "Logical", ",", "Philosophical", ",", "and", "Computational", "Foundations", "."], "sentence-detokenized": "M. Sowa a obtenu un BS en math\u00e9matiques au Massachusetts Institute of Technology en 1962, un MA en sciences appliqu\u00e9es \u00e0 l'Universit\u00e9 de Harvard en 1966 et un doctorat en informatique \u00e0 la Vrije Universiteit Brussel en 1999 sur une th\u00e8se intitul\u00e9e Knowledge Representation : Logical, Philosophical, and Computational Foundations.", "token2charspan": [[0, 2], [3, 7], [8, 9], [10, 16], [17, 19], [20, 22], [23, 25], [26, 39], [40, 42], [43, 56], [57, 66], [67, 69], [70, 80], [81, 83], [84, 88], [88, 89], [90, 92], [93, 95], [96, 98], [99, 107], [108, 118], [119, 120], [121, 123], [123, 133], [134, 136], [137, 144], [145, 147], [148, 152], [153, 155], [156, 158], [159, 167], [168, 170], [171, 183], [184, 185], [186, 188], [189, 194], [195, 207], [208, 215], [216, 218], [219, 223], [224, 227], [228, 231], [232, 237], [238, 247], [248, 257], [258, 272], [273, 274], [275, 282], [282, 283], [284, 297], [297, 298], [299, 302], [303, 316], [317, 328], [328, 329]]}
{"doc_key": "ai-dev-150", "ner": [[2, 4, "task"], [12, 12, "task"], [24, 24, "metrics"], [27, 28, "metrics"], [31, 32, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[2, 4, 12, 12, "general-affiliation", "", false, false], [24, 24, 2, 4, "part-of", "", true, false], [27, 28, 2, 4, "part-of", "", true, false], [31, 32, 2, 4, "part-of", "", true, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Puisque", "la", "reconnaissance", "des", "paraphrases", "peut", "\u00eatre", "pos\u00e9e", "comme", "un", "probl\u00e8me", "de", "classification", ",", "la", "plupart", "des", "mesures", "d'", "\u00e9valuation", "standard", "telles", "que", "la", "pr\u00e9cision", ",", "le", "score", "f1", "ou", "une", "courbe", "ROC", "donnent", "des", "r\u00e9sultats", "relativement", "bons", "."], "sentence-detokenized": "Puisque la reconnaissance des paraphrases peut \u00eatre pos\u00e9e comme un probl\u00e8me de classification, la plupart des mesures d'\u00e9valuation standard telles que la pr\u00e9cision, le score f1 ou une courbe ROC donnent des r\u00e9sultats relativement bons.", "token2charspan": [[0, 7], [8, 10], [11, 25], [26, 29], [30, 41], [42, 46], [47, 51], [52, 57], [58, 63], [64, 66], [67, 75], [76, 78], [79, 93], [93, 94], [95, 97], [98, 105], [106, 109], [110, 117], [118, 120], [120, 130], [131, 139], [140, 146], [147, 150], [151, 153], [154, 163], [163, 164], [165, 167], [168, 173], [174, 176], [177, 179], [180, 183], [184, 190], [191, 194], [195, 202], [203, 206], [207, 216], [217, 229], [230, 234], [234, 235]]}
{"doc_key": "ai-dev-151", "ner": [[24, 24, "algorithm"], [37, 39, "algorithm"], [42, 44, "algorithm"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[24, 24, 37, 39, "opposite", "not_suited_for", false, false], [24, 24, 42, 44, "opposite", "not_suited_for", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Cela", "la", "rend", "pratique", "pour", "l'", "analyse", "de", "grands", "ensembles", "de", "donn\u00e9es", "(", "des", "centaines", "ou", "des", "milliers", "de", "taxons", ")", "et", "pour", "le", "bootstrapping", ",", "pour", "lesquels", "d'", "autres", "moyens", "d'", "analyse", "(", "par", "exemple", "le", "maximum", "de", "parcimonie", ",", "le", "maximum", "de", "vraisemblance", ")", "peuvent", "\u00eatre", "prohibitifs", "en", "termes", "de", "calcul", "."], "sentence-detokenized": "Cela la rend pratique pour l'analyse de grands ensembles de donn\u00e9es (des centaines ou des milliers de taxons) et pour le bootstrapping, pour lesquels d'autres moyens d'analyse (par exemple le maximum de parcimonie, le maximum de vraisemblance) peuvent \u00eatre prohibitifs en termes de calcul.", "token2charspan": [[0, 4], [5, 7], [8, 12], [13, 21], [22, 26], [27, 29], [29, 36], [37, 39], [40, 46], [47, 56], [57, 59], [60, 67], [68, 69], [69, 72], [73, 82], [83, 85], [86, 89], [90, 98], [99, 101], [102, 108], [108, 109], [110, 112], [113, 117], [118, 120], [121, 134], [134, 135], [136, 140], [141, 149], [150, 152], [152, 158], [159, 165], [166, 168], [168, 175], [176, 177], [177, 180], [181, 188], [189, 191], [192, 199], [200, 202], [203, 213], [213, 214], [215, 217], [218, 225], [226, 228], [229, 242], [242, 243], [244, 251], [252, 256], [257, 268], [269, 271], [272, 278], [279, 281], [282, 288], [288, 289]]}
{"doc_key": "ai-dev-152", "ner": [[6, 6, "programlang"], [8, 8, "programlang"], [10, 13, "organisation"], [15, 15, "organisation"], [25, 25, "programlang"], [28, 39, "organisation"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[6, 6, 25, 25, "named", "same", false, false], [15, 15, 10, 13, "named", "", false, false], [28, 39, 6, 6, "role", "submits", true, false], [28, 39, 8, 8, "role", "submits", true, false], [28, 39, 10, 13, "role", "submits_to", true, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["La", "soumission", "en", "2002", "du", "langage", "DAML", "+", "OIL", "au", "World", "Wide", "Web", "Consortium", "(", "W3C", ")", ";", "le", "travail", "effectu\u00e9", "par", "les", "contractants", "de", "DAML", "et", "le", "comit\u00e9", "ad", "hoc", "Union", "europ\u00e9enne", "/", "\u00c9tats-Unis", "Joint", "Committee", "on", "Markup", "Languages", "."], "sentence-detokenized": "La soumission en 2002 du langage DAML + OIL au World Wide Web Consortium (W3C) ; le travail effectu\u00e9 par les contractants de DAML et le comit\u00e9 ad hoc Union europ\u00e9enne / \u00c9tats-Unis Joint Committee on Markup Languages.", "token2charspan": [[0, 2], [3, 13], [14, 16], [17, 21], [22, 24], [25, 32], [33, 37], [38, 39], [40, 43], [44, 46], [47, 52], [53, 57], [58, 61], [62, 72], [73, 74], [74, 77], [77, 78], [79, 80], [81, 83], [84, 91], [92, 100], [101, 104], [105, 108], [109, 121], [122, 124], [125, 129], [130, 132], [133, 135], [136, 142], [143, 145], [146, 149], [150, 155], [156, 166], [167, 168], [169, 179], [180, 185], [186, 195], [196, 198], [199, 205], [206, 215], [215, 216]]}
{"doc_key": "ai-dev-153", "ner": [[3, 5, "misc"], [10, 10, "misc"], [13, 14, "algorithm"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[10, 10, 3, 5, "part-of", "", true, false], [13, 14, 3, 5, "part-of", "", true, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Un", "exemple", "de", "normalisation", "non", "lin\u00e9aire", "est", "celui", "o\u00f9", "la", "normalisation", "suit", "une", "fonction", "sigmo\u00efde", ".", "Dans", "ce", "cas", ",", "l'", "image", "normalis\u00e9e", "est", "calcul\u00e9e", "selon", "la", "formule", "suivante"], "sentence-detokenized": "Un exemple de normalisation non lin\u00e9aire est celui o\u00f9 la normalisation suit une fonction sigmo\u00efde. Dans ce cas, l'image normalis\u00e9e est calcul\u00e9e selon la formule suivante", "token2charspan": [[0, 2], [3, 10], [11, 13], [14, 27], [28, 31], [32, 40], [41, 44], [45, 50], [51, 53], [54, 56], [57, 70], [71, 75], [76, 79], [80, 88], [89, 97], [97, 98], [99, 103], [104, 106], [107, 110], [110, 111], [112, 114], [114, 119], [120, 130], [131, 134], [135, 143], [144, 149], [150, 152], [153, 160], [161, 169]]}
{"doc_key": "ai-dev-154", "ner": [[6, 6, "metrics"], [11, 11, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[6, 6, 11, 11, "related-to", "used_together", false, false]], "relations_mapping_to_source": [0], "sentence": ["Il", "a", "\u00e9t\u00e9", "soulign\u00e9", "que", "la", "pr\u00e9cision", "est", "g\u00e9n\u00e9ralement", "jumel\u00e9e", "au", "rappel", "pour", "surmonter", "ce", "probl\u00e8me", "."], "sentence-detokenized": "Il a \u00e9t\u00e9 soulign\u00e9 que la pr\u00e9cision est g\u00e9n\u00e9ralement jumel\u00e9e au rappel pour surmonter ce probl\u00e8me.", "token2charspan": [[0, 2], [3, 4], [5, 8], [9, 17], [18, 21], [22, 24], [25, 34], [35, 38], [39, 51], [52, 59], [60, 62], [63, 69], [70, 74], [75, 84], [85, 87], [88, 96], [96, 97]]}
{"doc_key": "ai-dev-155", "ner": [[6, 8, "metrics"], [11, 13, "metrics"], [22, 23, "misc"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[22, 23, 11, 13, "usage", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Les", "mesures", "couramment", "utilis\u00e9es", "sont", "l'", "erreur", "quadratique", "moyenne", "et", "l'", "erreur", "quadratique", "moyenne", ",", "cette", "derni\u00e8re", "ayant", "\u00e9t\u00e9", "utilis\u00e9e", "dans", "le", "prix", "Netflix", "."], "sentence-detokenized": "Les mesures couramment utilis\u00e9es sont l'erreur quadratique moyenne et l'erreur quadratique moyenne, cette derni\u00e8re ayant \u00e9t\u00e9 utilis\u00e9e dans le prix Netflix.", "token2charspan": [[0, 3], [4, 11], [12, 22], [23, 32], [33, 37], [38, 40], [40, 46], [47, 58], [59, 66], [67, 69], [70, 72], [72, 78], [79, 90], [91, 98], [98, 99], [100, 105], [106, 114], [115, 120], [121, 124], [125, 133], [134, 138], [139, 141], [142, 146], [147, 154], [154, 155]]}
{"doc_key": "ai-dev-156", "ner": [[10, 12, "organisation"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["En", "ao\u00fbt", "2016", ",", "un", "programme", "de", "recherche", "avec", "l'", "University", "College", "Hospital", "a", "\u00e9t\u00e9", "annonc\u00e9", "dans", "le", "but", "de", "d\u00e9velopper", "un", "algorithme", "capable", "de", "diff\u00e9rencier", "automatiquement", "les", "tissus", "sains", "et", "canc\u00e9reux", "dans", "les", "zones", "de", "la", "t\u00eate", "et", "du", "cou", "."], "sentence-detokenized": "En ao\u00fbt 2016, un programme de recherche avec l'University College Hospital a \u00e9t\u00e9 annonc\u00e9 dans le but de d\u00e9velopper un algorithme capable de diff\u00e9rencier automatiquement les tissus sains et canc\u00e9reux dans les zones de la t\u00eate et du cou.", "token2charspan": [[0, 2], [3, 7], [8, 12], [12, 13], [14, 16], [17, 26], [27, 29], [30, 39], [40, 44], [45, 47], [47, 57], [58, 65], [66, 74], [75, 76], [77, 80], [81, 88], [89, 93], [94, 96], [97, 100], [101, 103], [104, 114], [115, 117], [118, 128], [129, 136], [137, 139], [140, 152], [153, 168], [169, 172], [173, 179], [180, 185], [186, 188], [189, 198], [199, 203], [204, 207], [208, 213], [214, 216], [217, 219], [220, 224], [225, 227], [228, 230], [231, 234], [234, 235]]}
{"doc_key": "ai-dev-157", "ner": [[8, 10, "researcher"], [21, 23, "organisation"], [27, 30, "organisation"], [34, 37, "organisation"], [41, 46, "organisation"], [50, 56, "organisation"], [60, 63, "organisation"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[8, 10, 21, 23, "role", "", false, false], [8, 10, 27, 30, "role", "", false, false], [8, 10, 34, 37, "role", "", false, false], [8, 10, 41, 46, "role", "", false, false], [8, 10, 50, 56, "role", "", false, false], [8, 10, 60, 63, "role", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["L'", "impact", "des", "contributions", "th\u00e9oriques", "et", "empiriques", "de", "Posner", "a", "\u00e9t\u00e9", "reconnu", "par", "l'", "octroi", "de", "bourses", "d'", "\u00e9tudes", "\u00e0", "l'", "American", "Psychological", "Association", ",", "\u00e0", "l'", "Association", "for", "Psychological", "Science", ",", "\u00e0", "la", "Society", "of", "Experimental", "Psychologists", ",", "\u00e0", "l'", "American", "Academy", "of", "Arts", "and", "Sciences", ",", "\u00e0", "l'", "American", "Association", "for", "the", "Advancement", "of", "Science", "et", "\u00e0", "la", "National", "Academy", "of", "Sciences", "."], "sentence-detokenized": "L'impact des contributions th\u00e9oriques et empiriques de Posner a \u00e9t\u00e9 reconnu par l'octroi de bourses d'\u00e9tudes \u00e0 l'American Psychological Association, \u00e0 l'Association for Psychological Science, \u00e0 la Society of Experimental Psychologists, \u00e0 l'American Academy of Arts and Sciences, \u00e0 l'American Association for the Advancement of Science et \u00e0 la National Academy of Sciences.", "token2charspan": [[0, 2], [2, 8], [9, 12], [13, 26], [27, 37], [38, 40], [41, 51], [52, 54], [55, 61], [62, 63], [64, 67], [68, 75], [76, 79], [80, 82], [82, 88], [89, 91], [92, 99], [100, 102], [102, 108], [109, 110], [111, 113], [113, 121], [122, 135], [136, 147], [147, 148], [149, 150], [151, 153], [153, 164], [165, 168], [169, 182], [183, 190], [190, 191], [192, 193], [194, 196], [197, 204], [205, 207], [208, 220], [221, 234], [234, 235], [236, 237], [238, 240], [240, 248], [249, 256], [257, 259], [260, 264], [265, 268], [269, 277], [277, 278], [279, 280], [281, 283], [283, 291], [292, 303], [304, 307], [308, 311], [312, 323], [324, 326], [327, 334], [335, 337], [338, 339], [340, 342], [343, 351], [352, 359], [360, 362], [363, 371], [371, 372]]}
{"doc_key": "ai-dev-158", "ner": [[1, 1, "product"], [9, 10, "field"], [14, 16, "task"], [19, 22, "task"], [24, 24, "task"], [28, 31, "task"], [33, 33, "task"], [37, 38, "field"], [41, 42, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "relations": [[1, 1, 9, 10, "usage", "", false, false], [14, 16, 9, 10, "part-of", "", false, false], [19, 22, 9, 10, "part-of", "", false, false], [24, 24, 19, 22, "named", "", false, false], [28, 31, 9, 10, "part-of", "", false, false], [33, 33, 28, 31, "named", "", false, false], [37, 38, 9, 10, "part-of", "", false, false], [41, 42, 9, 10, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "sentence": ["Ces", "chatbots", "intelligents", "font", "appel", "\u00e0", "toutes", "sortes", "d'", "intelligence", "artificielle", ",", "comme", "la", "mod\u00e9ration", "d'", "images", "et", "la", "compr\u00e9hension", "du", "langage", "naturel", "(", "NLU", ")", ",", "la", "g\u00e9n\u00e9ration", "de", "langage", "naturel", "(", "NLG", ")", ",", "l'", "apprentissage", "automatique", "et", "l'", "apprentissage", "profond", "."], "sentence-detokenized": "Ces chatbots intelligents font appel \u00e0 toutes sortes d'intelligence artificielle, comme la mod\u00e9ration d'images et la compr\u00e9hension du langage naturel (NLU), la g\u00e9n\u00e9ration de langage naturel (NLG), l'apprentissage automatique et l'apprentissage profond.", "token2charspan": [[0, 3], [4, 12], [13, 25], [26, 30], [31, 36], [37, 38], [39, 45], [46, 52], [53, 55], [55, 67], [68, 80], [80, 81], [82, 87], [88, 90], [91, 101], [102, 104], [104, 110], [111, 113], [114, 116], [117, 130], [131, 133], [134, 141], [142, 149], [150, 151], [151, 154], [154, 155], [155, 156], [157, 159], [160, 170], [171, 173], [174, 181], [182, 189], [190, 191], [191, 194], [194, 195], [195, 196], [197, 199], [199, 212], [213, 224], [225, 227], [228, 230], [230, 243], [244, 251], [251, 252]]}
{"doc_key": "ai-dev-159", "ner": [[6, 8, "metrics"], [10, 10, "metrics"], [13, 13, "metrics"], [16, 22, "metrics"], [29, 32, "metrics"], [34, 34, "metrics"], [37, 43, "metrics"], [48, 50, "metrics"], [52, 52, "metrics"], [55, 61, "metrics"], [68, 71, "metrics"], [73, 73, "metrics"], [76, 82, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12], "relations": [[10, 10, 6, 8, "named", "", false, false], [13, 13, 6, 8, "named", "", false, false], [16, 22, 6, 8, "named", "", false, false], [34, 34, 29, 32, "named", "", false, false], [37, 43, 29, 32, "named", "", false, false], [52, 52, 48, 50, "named", "", false, false], [55, 61, 48, 50, "named", "", false, false], [73, 73, 68, 71, "named", "", false, false], [76, 82, 68, 71, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "sentence": ["Les", "rapports", "de", "ligne", "sont", "la", "valeur", "pr\u00e9dictive", "positive", "(", "PPV", ",", "ou", "pr\u00e9cision", ")", "(", "TP", "/", "(", "TP", "+", "FP", ")", ")", ",", "avec", "en", "compl\u00e9ment", "le", "taux", "de", "fausses", "d\u00e9couvertes", "(", "FDR", ")", "(", "FP", "/", "(", "TP", "+", "FP", ")", ")", ";", "et", "la", "valeur", "pr\u00e9dictive", "n\u00e9gative", "(", "NPV", ")", "(", "TN", "/", "(", "TN", "+", "FN", ")", ")", ",", "avec", "en", "compl\u00e9ment", "le", "taux", "de", "fausses", "omissions", "(", "FOR", ")", "(", "FN", "/", "(", "TN", "+", "FN", ")", ")", "."], "sentence-detokenized": "Les rapports de ligne sont la valeur pr\u00e9dictive positive (PPV, ou pr\u00e9cision) (TP / (TP + FP)), avec en compl\u00e9ment le taux de fausses d\u00e9couvertes (FDR) (FP / (TP + FP)) ; et la valeur pr\u00e9dictive n\u00e9gative (NPV) (TN / (TN + FN)), avec en compl\u00e9ment le taux de fausses omissions (FOR) (FN / (TN + FN)).", "token2charspan": [[0, 3], [4, 12], [13, 15], [16, 21], [22, 26], [27, 29], [30, 36], [37, 47], [48, 56], [57, 58], [58, 61], [61, 62], [63, 65], [66, 75], [75, 76], [77, 78], [78, 80], [81, 82], [83, 84], [84, 86], [87, 88], [89, 91], [91, 92], [92, 93], [93, 94], [95, 99], [100, 102], [103, 113], [114, 116], [117, 121], [122, 124], [125, 132], [133, 144], [145, 146], [146, 149], [149, 150], [151, 152], [152, 154], [155, 156], [157, 158], [158, 160], [161, 162], [163, 165], [165, 166], [166, 167], [168, 169], [170, 172], [173, 175], [176, 182], [183, 193], [194, 202], [203, 204], [204, 207], [207, 208], [209, 210], [210, 212], [213, 214], [215, 216], [216, 218], [219, 220], [221, 223], [223, 224], [224, 225], [225, 226], [227, 231], [232, 234], [235, 245], [246, 248], [249, 253], [254, 256], [257, 264], [265, 274], [275, 276], [276, 279], [279, 280], [281, 282], [282, 284], [285, 286], [287, 288], [288, 290], [291, 292], [293, 295], [295, 296], [296, 297], [297, 298]]}
{"doc_key": "ai-dev-160", "ner": [[9, 9, "misc"], [17, 19, "algorithm"], [21, 21, "algorithm"], [26, 29, "algorithm"], [31, 31, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[21, 21, 17, 19, "named", "", false, false], [31, 31, 26, 29, "named", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Les", "informations", "sont", "un", "m\u00e9lange", "de", "sitemaps", "et", "de", "RSS", "et", "sont", "cr\u00e9\u00e9es", "\u00e0", "l'", "aide", "du", "mod\u00e8le", "d'", "information", "(", "IM", ")", "et", "de", "l'", "ontologie", "des", "ressources", "biom\u00e9dicales", "(", "BRO", ")", "."], "sentence-detokenized": "Les informations sont un m\u00e9lange de sitemaps et de RSS et sont cr\u00e9\u00e9es \u00e0 l'aide du mod\u00e8le d'information (IM) et de l'ontologie des ressources biom\u00e9dicales (BRO).", "token2charspan": [[0, 3], [4, 16], [17, 21], [22, 24], [25, 32], [33, 35], [36, 44], [45, 47], [48, 50], [51, 54], [55, 57], [58, 62], [63, 69], [70, 71], [72, 74], [74, 78], [79, 81], [82, 88], [89, 91], [91, 102], [103, 104], [104, 106], [106, 107], [108, 110], [111, 113], [114, 116], [116, 125], [126, 129], [130, 140], [141, 153], [154, 155], [155, 158], [158, 159], [159, 160]]}
{"doc_key": "ai-dev-161", "ner": [[0, 3, "task"], [9, 11, "algorithm"], [13, 16, "algorithm"], [23, 25, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[0, 3, 13, 16, "origin", "based_on", false, false], [13, 16, 9, 11, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["La", "reconnaissance", "de", "texte", "r\u00e9cente", "est", "bas\u00e9e", "sur", "un", "r\u00e9seau", "neuronal", "r\u00e9current", "(", "m\u00e9moire", "\u00e0", "long", "terme", ")", "et", "ne", "n\u00e9cessite", "pas", "de", "mod\u00e8le", "de", "langue", "."], "sentence-detokenized": "La reconnaissance de texte r\u00e9cente est bas\u00e9e sur un r\u00e9seau neuronal r\u00e9current (m\u00e9moire \u00e0 long terme) et ne n\u00e9cessite pas de mod\u00e8le de langue.", "token2charspan": [[0, 2], [3, 17], [18, 20], [21, 26], [27, 34], [35, 38], [39, 44], [45, 48], [49, 51], [52, 58], [59, 67], [68, 77], [78, 79], [79, 86], [87, 88], [89, 93], [94, 99], [99, 100], [101, 103], [104, 106], [107, 116], [117, 120], [121, 123], [124, 130], [131, 133], [134, 140], [140, 141]]}
{"doc_key": "ai-dev-162", "ner": [[1, 3, "misc"], [9, 10, "metrics"], [14, 15, "algorithm"], [19, 20, "metrics"], [24, 25, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[9, 10, 1, 3, "type-of", "", false, false], [14, 15, 9, 10, "related-to", "", true, false], [19, 20, 1, 3, "type-of", "", false, false], [24, 25, 19, 20, "related-to", "", true, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Les", "fonctions", "de", "perte", "les", "plus", "courantes", "sont", "la", "perte", "charni\u00e8re", "(", "pour", "les", "SVM", "lin\u00e9aires", ")", "et", "la", "perte", "logarithmique", "(", "pour", "la", "r\u00e9gression", "logistique", ")", "."], "sentence-detokenized": "Les fonctions de perte les plus courantes sont la perte charni\u00e8re (pour les SVM lin\u00e9aires) et la perte logarithmique (pour la r\u00e9gression logistique).", "token2charspan": [[0, 3], [4, 13], [14, 16], [17, 22], [23, 26], [27, 31], [32, 41], [42, 46], [47, 49], [50, 55], [56, 65], [66, 67], [67, 71], [72, 75], [76, 79], [80, 89], [89, 90], [91, 93], [94, 96], [97, 102], [103, 116], [117, 118], [118, 122], [123, 125], [126, 136], [137, 147], [147, 148], [148, 149]]}
{"doc_key": "ai-dev-163", "ner": [[0, 1, "metrics"], [12, 16, "metrics"], [18, 18, "metrics"], [22, 24, "metrics"], [26, 26, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[0, 1, 12, 16, "compare", "", false, false], [0, 1, 22, 24, "compare", "", false, false], [18, 18, 12, 16, "named", "", false, false], [26, 26, 22, 24, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Le", "SSIM", "est", "con\u00e7u", "pour", "am\u00e9liorer", "les", "m\u00e9thodes", "traditionnelles", "telles", "que", "le", "rapport", "signal", "/", "bruit", "maximal", "(", "PSNR", ")", "et", "l'", "erreur", "quadratique", "moyenne", "(", "MSE", ")", "."], "sentence-detokenized": "Le SSIM est con\u00e7u pour am\u00e9liorer les m\u00e9thodes traditionnelles telles que le rapport signal/bruit maximal (PSNR) et l'erreur quadratique moyenne (MSE).", "token2charspan": [[0, 2], [3, 7], [8, 11], [12, 17], [18, 22], [23, 32], [33, 36], [37, 45], [46, 61], [62, 68], [69, 72], [73, 75], [76, 83], [84, 90], [90, 91], [91, 96], [97, 104], [105, 106], [106, 110], [110, 111], [112, 114], [115, 117], [117, 123], [124, 135], [136, 143], [144, 145], [145, 148], [148, 149], [149, 150]]}
{"doc_key": "ai-dev-164", "ner": [[13, 14, "researcher"], [16, 17, "researcher"], [19, 20, "researcher"]], "ner_mapping_to_source": [0, 1, 2], "relations": [], "relations_mapping_to_source": [], "sentence": ["Ses", "travaux", "ont", "inspir\u00e9", "les", "g\u00e9n\u00e9rations", "suivantes", "de", "chercheurs", "en", "robotique", ",", "comme", "Rodney", "Brooks", ",", "Hans", "Moravec", "et", "Mark", "Tilden", "."], "sentence-detokenized": "Ses travaux ont inspir\u00e9 les g\u00e9n\u00e9rations suivantes de chercheurs en robotique, comme Rodney Brooks, Hans Moravec et Mark Tilden.", "token2charspan": [[0, 3], [4, 11], [12, 15], [16, 23], [24, 27], [28, 39], [40, 49], [50, 52], [53, 63], [64, 66], [67, 76], [76, 77], [78, 83], [84, 90], [91, 97], [97, 98], [99, 103], [104, 111], [112, 114], [115, 119], [120, 126], [126, 127]]}
{"doc_key": "ai-dev-165", "ner": [[22, 22, "algorithm"], [26, 28, "algorithm"]], "ner_mapping_to_source": [0, 1], "relations": [[26, 28, 22, 22, "origin", "based_on", false, false]], "relations_mapping_to_source": [0], "sentence": ["De", "plus", ",", "l'", "entra\u00eenement", "des", "impulsions", "n'", "est", "pas", "diff\u00e9rentiable", ",", "ce", "qui", "\u00e9limine", "les", "m\u00e9thodes", "d'", "entra\u00eenement", "bas\u00e9es", "sur", "la", "r\u00e9tropropagation", ",", "comme", "la", "descente", "de", "gradient", "."], "sentence-detokenized": "De plus, l'entra\u00eenement des impulsions n'est pas diff\u00e9rentiable, ce qui \u00e9limine les m\u00e9thodes d'entra\u00eenement bas\u00e9es sur la r\u00e9tropropagation, comme la descente de gradient.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 11], [11, 23], [24, 27], [28, 38], [39, 41], [41, 44], [45, 48], [49, 63], [63, 64], [65, 67], [68, 71], [72, 79], [80, 83], [84, 92], [93, 95], [95, 107], [108, 114], [115, 118], [119, 121], [122, 138], [138, 139], [140, 145], [146, 148], [149, 157], [158, 160], [161, 169], [169, 170]]}
{"doc_key": "ai-dev-166", "ner": [[8, 10, "metrics"], [17, 17, "metrics"], [22, 22, "task"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[8, 10, 17, 17, "related-to", "describes", false, false], [17, 17, 22, 22, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Ces", "relations", "peuvent", "\u00eatre", "facilement", "repr\u00e9sent\u00e9es", "par", "une", "matrice", "de", "confusion", ",", "un", "tableau", "qui", "d\u00e9crit", "la", "pr\u00e9cision", "d'", "un", "mod\u00e8le", "de", "classification", "."], "sentence-detokenized": "Ces relations peuvent \u00eatre facilement repr\u00e9sent\u00e9es par une matrice de confusion, un tableau qui d\u00e9crit la pr\u00e9cision d'un mod\u00e8le de classification.", "token2charspan": [[0, 3], [4, 13], [14, 21], [22, 26], [27, 37], [38, 50], [51, 54], [55, 58], [59, 66], [67, 69], [70, 79], [79, 80], [81, 83], [84, 91], [92, 95], [96, 102], [103, 105], [106, 115], [116, 118], [118, 120], [121, 127], [128, 130], [131, 145], [145, 146]]}
{"doc_key": "ai-dev-167", "ner": [[3, 13, "conference"], [15, 18, "conference"], [21, 21, "organisation"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[15, 18, 3, 13, "named", "", false, false], [21, 21, 3, 13, "physical", "", false, false], [21, 21, 3, 13, "role", "", false, false], [21, 21, 3, 13, "temporal", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Lors", "de", "la", "conf\u00e9rence", "2018", "sur", "les", "syst\u00e8mes", "de", "traitement", "de", "l'", "information", "neuronale", "(", "NeurIPS", ")", ",", "des", "chercheurs", "de", "Google", "ont", "pr\u00e9sent\u00e9", "les", "travaux", "."], "sentence-detokenized": "Lors de la conf\u00e9rence 2018 sur les syst\u00e8mes de traitement de l'information neuronale (NeurIPS), des chercheurs de Google ont pr\u00e9sent\u00e9 les travaux.", "token2charspan": [[0, 4], [5, 7], [8, 10], [11, 21], [22, 26], [27, 30], [31, 34], [35, 43], [44, 46], [47, 57], [58, 60], [61, 63], [63, 74], [75, 84], [85, 86], [86, 93], [93, 94], [94, 95], [96, 99], [100, 110], [111, 113], [114, 120], [121, 124], [125, 133], [134, 137], [138, 145], [145, 146]]}
{"doc_key": "ai-dev-168", "ner": [[4, 4, "university"], [17, 17, "product"], [23, 26, "misc"], [31, 31, "conference"], [37, 40, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[17, 17, 23, 26, "win-defeat", "", false, false], [23, 26, 31, 31, "temporal", "", false, false], [37, 40, 31, 31, "part-of", "", false, false], [37, 40, 31, 31, "temporal", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Pendant", "son", "s\u00e9jour", "\u00e0", "Duke", ",", "il", "a", "travaill\u00e9", "sur", "un", "r\u00e9solveur", "automatique", "de", "mots", "crois\u00e9s", ",", "PROVERB", ",", "qui", "a", "remport\u00e9", "le", "prix", "Outstanding", "Paper", "Award", "en", "1999", "de", "l'", "AAAI", "et", "a", "particip\u00e9", "\u00e0", "l'", "American", "Crossword", "Puzzle", "Tournament", "."], "sentence-detokenized": "Pendant son s\u00e9jour \u00e0 Duke, il a travaill\u00e9 sur un r\u00e9solveur automatique de mots crois\u00e9s, PROVERB, qui a remport\u00e9 le prix Outstanding Paper Award en 1999 de l'AAAI et a particip\u00e9 \u00e0 l'American Crossword Puzzle Tournament.", "token2charspan": [[0, 7], [8, 11], [12, 18], [19, 20], [21, 25], [25, 26], [27, 29], [30, 31], [32, 41], [42, 45], [46, 48], [49, 58], [59, 70], [71, 73], [74, 78], [79, 86], [86, 87], [88, 95], [95, 96], [97, 100], [101, 102], [103, 111], [112, 114], [115, 119], [120, 131], [132, 137], [138, 143], [144, 146], [147, 151], [152, 154], [155, 157], [157, 161], [162, 164], [165, 166], [167, 176], [177, 178], [179, 181], [181, 189], [190, 199], [200, 206], [207, 217], [217, 218]]}
{"doc_key": "ai-dev-169", "ner": [[6, 7, "location"], [11, 11, "location"], [21, 21, "country"], [23, 24, "country"], [27, 27, "country"], [30, 30, "country"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[6, 7, 11, 11, "physical", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Le", "si\u00e8ge", "social", "se", "trouve", "\u00e0", "Rochester", "Hills", ",", "dans", "le", "Michigan", ",", "et", "la", "soci\u00e9t\u00e9", "poss\u00e8de", "10", "sites", "r\u00e9gionaux", "aux", "\u00c9tats-Unis", ",", "au", "Canada", ",", "au", "Mexique", "et", "au", "Br\u00e9sil", "."], "sentence-detokenized": "Le si\u00e8ge social se trouve \u00e0 Rochester Hills, dans le Michigan, et la soci\u00e9t\u00e9 poss\u00e8de 10 sites r\u00e9gionaux aux \u00c9tats-Unis, au Canada, au Mexique et au Br\u00e9sil.", "token2charspan": [[0, 2], [3, 8], [9, 15], [16, 18], [19, 25], [26, 27], [28, 37], [38, 43], [43, 44], [45, 49], [50, 52], [53, 61], [61, 62], [63, 65], [66, 68], [69, 76], [77, 84], [85, 87], [88, 93], [94, 103], [104, 107], [108, 118], [118, 119], [120, 122], [123, 129], [129, 130], [131, 133], [134, 141], [142, 144], [145, 147], [148, 154], [154, 155]]}
{"doc_key": "ai-dev-170", "ner": [[12, 12, "product"], [15, 17, "product"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "rejoint", "une", "collection", "de", "robots", "historiquement", "importants", "qui", "comprend", "un", "ancien", "Unimate", "et", "l'", "Odetics", "Odex", "1", "."], "sentence-detokenized": "Il rejoint une collection de robots historiquement importants qui comprend un ancien Unimate et l'Odetics Odex 1.", "token2charspan": [[0, 2], [3, 10], [11, 14], [15, 25], [26, 28], [29, 35], [36, 50], [51, 61], [62, 65], [66, 74], [75, 77], [78, 84], [85, 92], [93, 95], [96, 98], [98, 105], [106, 110], [111, 112], [112, 113]]}
{"doc_key": "ai-dev-171", "ner": [[11, 11, "researcher"], [13, 13, "organisation"], [15, 16, "researcher"], [26, 30, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[11, 11, 13, 13, "physical", "", false, false], [11, 11, 13, 13, "role", "", false, false], [15, 16, 13, 13, "physical", "", false, false], [15, 16, 13, 13, "role", "", false, false], [15, 16, 26, 30, "win-defeat", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["L'", "\u00e9diteur", "invit\u00e9", "pour", "ce", "num\u00e9ro", "sera", "l'", "ancien", "coll\u00e8gue", "de", "David", "au", "NIST", ",", "Judah", "Levine", ",", "qui", "est", "le", "plus", "r\u00e9cent", "laur\u00e9at", "du", "prix", "I.", "I", "..", "Rabi", "Award", "."], "sentence-detokenized": "L'\u00e9diteur invit\u00e9 pour ce num\u00e9ro sera l'ancien coll\u00e8gue de David au NIST, Judah Levine, qui est le plus r\u00e9cent laur\u00e9at du prix I. I.. Rabi Award.", "token2charspan": [[0, 2], [2, 9], [10, 16], [17, 21], [22, 24], [25, 31], [32, 36], [37, 39], [39, 45], [46, 54], [55, 57], [58, 63], [64, 66], [67, 71], [71, 72], [73, 78], [79, 85], [85, 86], [87, 90], [91, 94], [95, 97], [98, 102], [103, 109], [110, 117], [118, 120], [121, 125], [126, 128], [129, 130], [130, 132], [133, 137], [138, 143], [143, 144]]}
{"doc_key": "ai-dev-172", "ner": [[13, 15, "metrics"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Ceux-ci", "peuvent", "\u00eatre", "organis\u00e9s", "en", "un", "tableau", "de", "contingence", "2", "\u00d7", "2", "(", "matrice", "de", "confusion", ")", ",", "avec", "par", "convention", "le", "r\u00e9sultat", "du", "test", "sur", "l'", "axe", "vertical", "et", "la", "condition", "r\u00e9elle", "sur", "l'", "axe", "horizontal", "."], "sentence-detokenized": "Ceux-ci peuvent \u00eatre organis\u00e9s en un tableau de contingence 2 \u00d7 2 (matrice de confusion), avec par convention le r\u00e9sultat du test sur l'axe vertical et la condition r\u00e9elle sur l'axe horizontal.", "token2charspan": [[0, 7], [8, 15], [16, 20], [21, 30], [31, 33], [34, 36], [37, 44], [45, 47], [48, 59], [60, 61], [62, 63], [64, 65], [66, 67], [67, 74], [75, 77], [78, 87], [87, 88], [88, 89], [90, 94], [95, 98], [99, 109], [110, 112], [113, 121], [122, 124], [125, 129], [130, 133], [134, 136], [136, 139], [140, 148], [149, 151], [152, 154], [155, 164], [165, 171], [172, 175], [176, 178], [178, 181], [182, 192], [192, 193]]}
{"doc_key": "ai-dev-173", "ner": [[1, 5, "product"], [9, 9, "product"], [12, 12, "product"], [15, 16, "product"], [21, 23, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[1, 5, 9, 9, "part-of", "", false, false], [1, 5, 12, 12, "part-of", "", false, false], [1, 5, 15, 16, "part-of", "", false, false], [1, 5, 21, 23, "usage", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Le", "syst\u00e8me", "d'", "exploitation", "Apple", "iOS", "utilis\u00e9", "sur", "l'", "iPhone", ",", "l'", "iPad", "et", "l'", "iPod", "Touch", "utilise", "l'", "accessibilit\u00e9", "par", "synth\u00e8se", "vocale", "VoiceOver", "."], "sentence-detokenized": "Le syst\u00e8me d'exploitation Apple iOS utilis\u00e9 sur l'iPhone, l'iPad et l'iPod Touch utilise l'accessibilit\u00e9 par synth\u00e8se vocale VoiceOver.", "token2charspan": [[0, 2], [3, 10], [11, 13], [13, 25], [26, 31], [32, 35], [36, 43], [44, 47], [48, 50], [50, 56], [56, 57], [58, 60], [60, 64], [65, 67], [68, 70], [70, 74], [75, 80], [81, 88], [89, 91], [91, 104], [105, 108], [109, 117], [118, 124], [125, 134], [134, 135]]}
{"doc_key": "ai-dev-174", "ner": [[8, 8, "conference"], [14, 14, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Par", "exemple", ",", "le", "meilleur", "syst\u00e8me", "entrant", "dans", "MUC-7", "a", "obtenu", "93,39", "%", "de", "F-mesure", "alors", "que", "les", "annotateurs", "humains", "ont", "obtenu", "97,6", "%", "et", "96,95", "%", "."], "sentence-detokenized": "Par exemple, le meilleur syst\u00e8me entrant dans MUC-7 a obtenu 93,39% de F-mesure alors que les annotateurs humains ont obtenu 97,6% et 96,95%.", "token2charspan": [[0, 3], [4, 11], [11, 12], [13, 15], [16, 24], [25, 32], [33, 40], [41, 45], [46, 51], [52, 53], [54, 60], [61, 66], [66, 67], [68, 70], [71, 79], [80, 85], [86, 89], [90, 93], [94, 105], [106, 113], [114, 117], [118, 124], [125, 129], [129, 130], [131, 133], [134, 139], [139, 140], [140, 141]]}
{"doc_key": "ai-dev-175", "ner": [[19, 22, "algorithm"], [24, 24, "algorithm"]], "ner_mapping_to_source": [0, 1], "relations": [[24, 24, 19, 22, "part-of", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Cette", "op\u00e9ration", "s'", "effectue", "\u00e0", "l'", "aide", "d'", "algorithmes", "standard", "de", "formation", "de", "r\u00e9seaux", "neuronaux", ",", "tels", "que", "la", "descente", "de", "gradient", "stochastique", "avec", "r\u00e9tropropagation", "."], "sentence-detokenized": "Cette op\u00e9ration s'effectue \u00e0 l'aide d'algorithmes standard de formation de r\u00e9seaux neuronaux, tels que la descente de gradient stochastique avec r\u00e9tropropagation.", "token2charspan": [[0, 5], [6, 15], [16, 18], [18, 26], [27, 28], [29, 31], [31, 35], [36, 38], [38, 49], [50, 58], [59, 61], [62, 71], [72, 74], [75, 82], [83, 92], [92, 93], [94, 98], [99, 102], [103, 105], [106, 114], [115, 117], [118, 126], [127, 139], [140, 144], [145, 161], [161, 162]]}
{"doc_key": "ai-dev-176", "ner": [[0, 1, "organisation"], [24, 24, "country"], [32, 32, "product"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[0, 1, 24, 24, "physical", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Rotten", "Tomatoes", "est", "un", "site", "du", "top", "1000", ",", "qui", "se", "situe", "aux", "alentours", "du", "400e", "rang", "mondial", "et", "du", "150e", "rang", "pour", "les", "\u00c9tats-Unis", "uniquement", ",", "selon", "le", "site", "de", "classement", "Alexa", "."], "sentence-detokenized": "Rotten Tomatoes est un site du top 1000, qui se situe aux alentours du 400e rang mondial et du 150e rang pour les \u00c9tats-Unis uniquement, selon le site de classement Alexa.", "token2charspan": [[0, 6], [7, 15], [16, 19], [20, 22], [23, 27], [28, 30], [31, 34], [35, 39], [39, 40], [41, 44], [45, 47], [48, 53], [54, 57], [58, 67], [68, 70], [71, 75], [76, 80], [81, 88], [89, 91], [92, 94], [95, 99], [100, 104], [105, 109], [110, 113], [114, 124], [125, 135], [135, 136], [137, 142], [143, 145], [146, 150], [151, 153], [154, 164], [165, 170], [170, 171]]}
{"doc_key": "ai-dev-177", "ner": [[18, 19, "algorithm"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["D'", "une", "mani\u00e8re", "g\u00e9n\u00e9rale", ",", "tout", "apprentissage", "pr\u00e9sente", "une", "\u00e9volution", "progressive", "dans", "le", "temps", ",", "mais", "d\u00e9crit", "une", "fonction", "sigmo\u00efde", "qui", "a", "des", "apparences", "diff\u00e9rentes", "selon", "l'", "\u00e9chelle", "de", "temps", "d'", "observation", "."], "sentence-detokenized": "D'une mani\u00e8re g\u00e9n\u00e9rale, tout apprentissage pr\u00e9sente une \u00e9volution progressive dans le temps, mais d\u00e9crit une fonction sigmo\u00efde qui a des apparences diff\u00e9rentes selon l'\u00e9chelle de temps d'observation.", "token2charspan": [[0, 2], [2, 5], [6, 13], [14, 22], [22, 23], [24, 28], [29, 42], [43, 51], [52, 55], [56, 65], [66, 77], [78, 82], [83, 85], [86, 91], [91, 92], [93, 97], [98, 104], [105, 108], [109, 117], [118, 126], [127, 130], [131, 132], [133, 136], [137, 147], [148, 159], [160, 165], [166, 168], [168, 175], [176, 178], [179, 184], [185, 187], [187, 198], [198, 199]]}
{"doc_key": "ai-dev-178", "ner": [[1, 1, "metrics"], [9, 11, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[1, 1, 9, 11, "named", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["La", "DSS", "est", "\u00e9galement", "connue", "sous", "le", "nom", "d'", "erreur", "quadratique", "moyenne", "."], "sentence-detokenized": "La DSS est \u00e9galement connue sous le nom d'erreur quadratique moyenne.", "token2charspan": [[0, 2], [3, 6], [7, 10], [11, 20], [21, 27], [28, 32], [33, 35], [36, 39], [40, 42], [42, 48], [49, 60], [61, 68], [68, 69]]}
{"doc_key": "ai-dev-179", "ner": [[0, 5, "algorithm"], [8, 9, "algorithm"], [12, 14, "algorithm"], [31, 32, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[0, 5, 31, 32, "related-to", "can_be_related_to", true, false], [8, 9, 31, 32, "related-to", "can_be_related_to", true, false], [12, 14, 31, 32, "related-to", "can_be_related_to", true, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["L'", "apprentissage", "par", "arbre", "de", "d\u00e9cision", ",", "les", "r\u00e9seaux", "neuronaux", "ou", "un", "classificateur", "Bayes", "na\u00eff", "pourraient", "\u00eatre", "utilis\u00e9s", "en", "combinaison", "avec", "des", "mesures", "de", "la", "qualit\u00e9", "du", "mod\u00e8le", "telles", "que", "la", "pr\u00e9cision", "\u00e9quilibr\u00e9e"], "sentence-detokenized": "L'apprentissage par arbre de d\u00e9cision, les r\u00e9seaux neuronaux ou un classificateur Bayes na\u00eff pourraient \u00eatre utilis\u00e9s en combinaison avec des mesures de la qualit\u00e9 du mod\u00e8le telles que la pr\u00e9cision \u00e9quilibr\u00e9e", "token2charspan": [[0, 2], [2, 15], [16, 19], [20, 25], [26, 28], [29, 37], [37, 38], [39, 42], [43, 50], [51, 60], [61, 63], [64, 66], [67, 81], [82, 87], [88, 92], [93, 103], [104, 108], [109, 117], [118, 120], [121, 132], [133, 137], [138, 141], [142, 149], [150, 152], [153, 155], [156, 163], [164, 166], [167, 173], [174, 180], [181, 184], [185, 187], [188, 197], [198, 208]]}
{"doc_key": "ai-dev-180", "ner": [[17, 17, "conference"], [23, 29, "conference"], [22, 31, "misc"], [36, 39, "product"], [46, 49, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[22, 31, 23, 29, "origin", "", false, false], [22, 31, 23, 29, "temporal", "", false, false], [36, 39, 22, 31, "win-defeat", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Il", "est", "un", "ancien", "pr\u00e9sident", "(", "1979", ")", "et", "un", "membre", "inaugural", "(", "2011", ")", "de", "l'", "ACL", ",", "un", "cor\u00e9cipiendaire", "du", "prix", "1992", "de", "l'", "Association", "for", "Computing", "Machinery", "Software", "Systems", "pour", "sa", "contribution", "au", "syst\u00e8me", "de", "programmation", "Interlisp", ",", "et", "un", "membre", "de", "l'", "Association", "for", "Computing", "Machinery", "."], "sentence-detokenized": "Il est un ancien pr\u00e9sident (1979) et un membre inaugural (2011) de l'ACL, un cor\u00e9cipiendaire du prix 1992 de l'Association for Computing Machinery Software Systems pour sa contribution au syst\u00e8me de programmation Interlisp, et un membre de l'Association for Computing Machinery.", "token2charspan": [[0, 2], [3, 6], [7, 9], [10, 16], [17, 26], [27, 28], [28, 32], [32, 33], [34, 36], [37, 39], [40, 46], [47, 56], [57, 58], [58, 62], [62, 63], [64, 66], [67, 69], [69, 72], [72, 73], [74, 76], [77, 92], [93, 95], [96, 100], [101, 105], [106, 108], [109, 111], [111, 122], [123, 126], [127, 136], [137, 146], [147, 155], [156, 163], [164, 168], [169, 171], [172, 184], [185, 187], [188, 195], [196, 198], [199, 212], [213, 222], [222, 223], [224, 226], [227, 229], [230, 236], [237, 239], [240, 242], [242, 253], [254, 257], [258, 267], [268, 277], [277, 278]]}
{"doc_key": "ai-dev-181", "ner": [[1, 2, "researcher"], [4, 5, "researcher"], [7, 7, "researcher"], [11, 14, "researcher"], [27, 28, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[1, 2, 27, 28, "related-to", "", false, false], [4, 5, 27, 28, "related-to", "", false, false], [7, 7, 27, 28, "related-to", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Avec", "Geoffrey", "Hinton", "et", "Yann", "LeCun", ",", "Bengio", "est", "consid\u00e9r\u00e9", "par", "Cade", "Metz", "comme", "l'", "une", "des", "trois", "personnes", "les", "plus", "responsables", "de", "l'", "avancement", "de", "l'", "apprentissage", "profond", "dans", "les", "ann\u00e9es", "1990", "et", "2000", "."], "sentence-detokenized": "Avec Geoffrey Hinton et Yann LeCun, Bengio est consid\u00e9r\u00e9 par Cade Metz comme l'une des trois personnes les plus responsables de l'avancement de l'apprentissage profond dans les ann\u00e9es 1990 et 2000.", "token2charspan": [[0, 4], [5, 13], [14, 20], [21, 23], [24, 28], [29, 34], [34, 35], [36, 42], [43, 46], [47, 56], [57, 60], [61, 65], [66, 70], [71, 76], [77, 79], [79, 82], [83, 86], [87, 92], [93, 102], [103, 106], [107, 111], [112, 124], [125, 127], [128, 130], [130, 140], [141, 143], [144, 146], [146, 159], [160, 167], [168, 172], [173, 176], [177, 183], [184, 188], [189, 191], [192, 196], [196, 197]]}
{"doc_key": "ai-dev-182", "ner": [[1, 4, "field"], [7, 7, "field"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["En", "th\u00e9orie", "de", "l'", "information", "et", "en", "informatique", ",", "un", "code", "est", "g\u00e9n\u00e9ralement", "consid\u00e9r\u00e9", "comme", "un", "algorithme", "qui", "repr\u00e9sente", "de", "mani\u00e8re", "unique", "des", "symboles", "d'", "un", "certain", "alphabet", "source", ",", "par", "des", "cha\u00eenes", "cod\u00e9es", ",", "qui", "peuvent", "\u00eatre", "dans", "un", "autre", "alphabet", "cible", "."], "sentence-detokenized": "En th\u00e9orie de l'information et en informatique, un code est g\u00e9n\u00e9ralement consid\u00e9r\u00e9 comme un algorithme qui repr\u00e9sente de mani\u00e8re unique des symboles d'un certain alphabet source, par des cha\u00eenes cod\u00e9es, qui peuvent \u00eatre dans un autre alphabet cible.", "token2charspan": [[0, 2], [3, 10], [11, 13], [14, 16], [16, 27], [28, 30], [31, 33], [34, 46], [46, 47], [48, 50], [51, 55], [56, 59], [60, 72], [73, 82], [83, 88], [89, 91], [92, 102], [103, 106], [107, 117], [118, 120], [121, 128], [129, 135], [136, 139], [140, 148], [149, 151], [151, 153], [154, 161], [162, 170], [171, 177], [177, 178], [179, 182], [183, 186], [187, 194], [195, 201], [201, 202], [203, 206], [207, 214], [215, 219], [220, 224], [225, 227], [228, 233], [234, 242], [243, 248], [248, 249]]}
{"doc_key": "ai-dev-183", "ner": [[7, 8, "algorithm"], [12, 13, "algorithm"]], "ner_mapping_to_source": [0, 1], "relations": [[12, 13, 7, 8, "type-of", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Fonction", "non", "lin\u00e9aire", "assez", "simple", ",", "la", "fonction", "sigmo\u00efde", ",", "comme", "la", "fonction", "logistique", ",", "a", "\u00e9galement", "une", "d\u00e9riv\u00e9e", "facile", "\u00e0", "calculer", ",", "ce", "qui", "peut", "\u00eatre", "important", "lors", "du", "calcul", "des", "mises", "\u00e0", "jour", "des", "poids", "dans", "le", "r\u00e9seau", "."], "sentence-detokenized": "Fonction non lin\u00e9aire assez simple, la fonction sigmo\u00efde, comme la fonction logistique, a \u00e9galement une d\u00e9riv\u00e9e facile \u00e0 calculer, ce qui peut \u00eatre important lors du calcul des mises \u00e0 jour des poids dans le r\u00e9seau.", "token2charspan": [[0, 8], [9, 12], [13, 21], [22, 27], [28, 34], [34, 35], [36, 38], [39, 47], [48, 56], [56, 57], [58, 63], [64, 66], [67, 75], [76, 86], [86, 87], [88, 89], [90, 99], [100, 103], [104, 111], [112, 118], [119, 120], [121, 129], [129, 130], [131, 133], [134, 137], [138, 142], [143, 147], [148, 157], [158, 162], [163, 165], [166, 172], [173, 176], [177, 182], [183, 184], [185, 189], [190, 193], [194, 199], [200, 204], [205, 207], [208, 214], [214, 215]]}
{"doc_key": "ai-dev-184", "ner": [[0, 0, "person"], [4, 4, "location"], [7, 7, "location"], [9, 9, "country"], [12, 12, "country"], [15, 16, "country"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[0, 0, 4, 4, "physical", "", false, false], [4, 4, 7, 7, "physical", "", false, false], [7, 7, 9, 9, "physical", "", false, false], [7, 7, 12, 12, "physical", "", false, false], [7, 7, 15, 16, "physical", "", false, false], [12, 12, 9, 9, "origin", "", false, false], [15, 16, 12, 12, "origin", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "sentence": ["\u010capek", "est", "n\u00e9", "\u00e0", "Hronov", ",", "en", "Boh\u00eame", "(", "Autriche-Hongrie", ",", "puis", "Tch\u00e9coslovaquie", ",", "aujourd'hui", "R\u00e9publique", "tch\u00e8que", ")", "en", "1887", "."], "sentence-detokenized": "\u010capek est n\u00e9 \u00e0 Hronov, en Boh\u00eame (Autriche-Hongrie, puis Tch\u00e9coslovaquie, aujourd'hui R\u00e9publique tch\u00e8que) en 1887.", "token2charspan": [[0, 5], [6, 9], [10, 12], [13, 14], [15, 21], [21, 22], [23, 25], [26, 32], [33, 34], [34, 50], [50, 51], [52, 56], [57, 72], [72, 73], [74, 85], [86, 96], [97, 104], [104, 105], [106, 108], [109, 113], [113, 114]]}
{"doc_key": "ai-dev-185", "ner": [[6, 6, "product"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Certains", "logiciels", "sp\u00e9cialis\u00e9s", "peuvent", "narrer", "des", "RSS", "."], "sentence-detokenized": "Certains logiciels sp\u00e9cialis\u00e9s peuvent narrer des RSS.", "token2charspan": [[0, 8], [9, 18], [19, 30], [31, 38], [39, 45], [46, 49], [50, 53], [53, 54]]}
{"doc_key": "ai-dev-186", "ner": [[13, 14, "task"], [17, 19, "task"], [24, 24, "task"], [27, 27, "task"], [30, 32, "task"], [43, 45, "task"], [48, 52, "task"], [58, 58, "task"], [61, 61, "product"], [63, 64, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9], "relations": [[13, 14, 17, 19, "related-to", "", true, false], [13, 14, 24, 24, "related-to", "", true, false], [13, 14, 27, 27, "related-to", "", true, false], [48, 52, 43, 45, "usage", "", true, false], [61, 61, 58, 58, "type-of", "", false, false], [63, 64, 58, 58, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["Parmi", "les", "aspects", "des", "\u00e9diteurs", "d'", "ontologie", ",", "citons", ":", "les", "possibilit\u00e9s", "de", "navigation", "visuelle", "dans", "le", "mod\u00e8le", "de", "connaissances", ",", "les", "moteurs", "d'", "inf\u00e9rence", "et", "l'", "extraction", ";", "le", "support", "des", "modules", ";", "l'", "importation", "et", "l'", "exportation", "de", "langages", "\u00e9trangers", "de", "repr\u00e9sentation", "des", "connaissances", "pour", "la", "mise", "en", "correspondance", "des", "ontologies", ";", "et", "le", "support", "des", "m\u00e9ta-ontologies", "telles", "que", "OWL-S", ",", "Dublin", "Core", ",", "etc."], "sentence-detokenized": "Parmi les aspects des \u00e9diteurs d'ontologie, citons : les possibilit\u00e9s de navigation visuelle dans le mod\u00e8le de connaissances, les moteurs d'inf\u00e9rence et l'extraction ; le support des modules ; l'importation et l'exportation de langages \u00e9trangers de repr\u00e9sentation des connaissances pour la mise en correspondance des ontologies ; et le support des m\u00e9ta-ontologies telles que OWL-S, Dublin Core, etc.", "token2charspan": [[0, 5], [6, 9], [10, 17], [18, 21], [22, 30], [31, 33], [33, 42], [42, 43], [44, 50], [51, 52], [53, 56], [57, 69], [70, 72], [73, 83], [84, 92], [93, 97], [98, 100], [101, 107], [108, 110], [111, 124], [124, 125], [126, 129], [130, 137], [138, 140], [140, 149], [150, 152], [153, 155], [155, 165], [166, 167], [168, 170], [171, 178], [179, 182], [183, 190], [191, 192], [193, 195], [195, 206], [207, 209], [210, 212], [212, 223], [224, 226], [227, 235], [236, 245], [246, 248], [249, 263], [264, 267], [268, 281], [282, 286], [287, 289], [290, 294], [295, 297], [298, 312], [313, 316], [317, 327], [328, 329], [330, 332], [333, 335], [336, 343], [344, 347], [348, 363], [364, 370], [371, 374], [375, 380], [380, 381], [382, 388], [389, 393], [393, 394], [395, 399]]}
{"doc_key": "ai-dev-187", "ner": [[1, 1, "organisation"], [8, 13, "misc"], [18, 19, "task"], [24, 25, "field"], [30, 31, "misc"], [34, 37, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[8, 13, 1, 1, "origin", "", false, false], [18, 19, 8, 13, "part-of", "", false, false], [24, 25, 8, 13, "part-of", "", false, false], [30, 31, 24, 25, "type-of", "", false, false], [34, 37, 24, 25, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Le", "FBI", "a", "\u00e9galement", "mis", "en", "place", "son", "programme", "d'", "identification", "de", "nouvelle", "g\u00e9n\u00e9ration", ",", "qui", "inclut", "la", "reconnaissance", "faciale", ",", "ainsi", "que", "des", "\u00e9l\u00e9ments", "biom\u00e9triques", "plus", "traditionnels", "comme", "les", "empreintes", "digitales", "et", "les", "scans", "de", "l'", "iris", ",", "qui", "peuvent", "\u00eatre", "extraits", "de", "bases", "de", "donn\u00e9es", "criminelles", "et", "civiles", "."], "sentence-detokenized": "Le FBI a \u00e9galement mis en place son programme d'identification de nouvelle g\u00e9n\u00e9ration, qui inclut la reconnaissance faciale, ainsi que des \u00e9l\u00e9ments biom\u00e9triques plus traditionnels comme les empreintes digitales et les scans de l'iris, qui peuvent \u00eatre extraits de bases de donn\u00e9es criminelles et civiles.", "token2charspan": [[0, 2], [3, 6], [7, 8], [9, 18], [19, 22], [23, 25], [26, 31], [32, 35], [36, 45], [46, 48], [48, 62], [63, 65], [66, 74], [75, 85], [85, 86], [87, 90], [91, 97], [98, 100], [101, 115], [116, 123], [123, 124], [125, 130], [131, 134], [135, 138], [139, 147], [148, 160], [161, 165], [166, 179], [180, 185], [186, 189], [190, 200], [201, 210], [211, 213], [214, 217], [218, 223], [224, 226], [227, 229], [229, 233], [233, 234], [235, 238], [239, 246], [247, 251], [252, 260], [261, 263], [264, 269], [270, 272], [273, 280], [281, 292], [293, 295], [296, 303], [303, 304]]}
{"doc_key": "ai-dev-188", "ner": [[5, 7, "person"], [16, 17, "person"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Pour", "la", "saison", "2016", ",", "Samantha", "Ponder", "a", "\u00e9t\u00e9", "ajout\u00e9e", "comme", "animatrice", ",", "en", "remplacement", "de", "Molly", "McGrath", "."], "sentence-detokenized": "Pour la saison 2016, Samantha Ponder a \u00e9t\u00e9 ajout\u00e9e comme animatrice, en remplacement de Molly McGrath.", "token2charspan": [[0, 4], [5, 7], [8, 14], [15, 19], [19, 20], [21, 29], [30, 36], [37, 38], [39, 42], [43, 50], [51, 56], [57, 67], [67, 68], [69, 71], [72, 84], [85, 87], [88, 93], [94, 101], [101, 102]]}
{"doc_key": "ai-dev-189", "ner": [[5, 8, "algorithm"], [21, 22, "misc"], [24, 24, "misc"], [26, 26, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "s'", "agit", "d'", "un", "algorithme", "de", "recherche", "contradictoire", "utilis\u00e9", "couramment", "pour", "le", "jeu", "automatique", "de", "jeux", "\u00e0", "deux", "joueurs", "(", "Tic-tac", "-toe", ",", "\u00e9checs", ",", "go", ",", "etc.", ")", "."], "sentence-detokenized": "Il s'agit d'un algorithme de recherche contradictoire utilis\u00e9 couramment pour le jeu automatique de jeux \u00e0 deux joueurs (Tic-tac-toe, \u00e9checs, go, etc.).", "token2charspan": [[0, 2], [3, 5], [5, 9], [10, 12], [12, 14], [15, 25], [26, 28], [29, 38], [39, 53], [54, 61], [62, 72], [73, 77], [78, 80], [81, 84], [85, 96], [97, 99], [100, 104], [105, 106], [107, 111], [112, 119], [120, 121], [121, 128], [128, 132], [132, 133], [134, 140], [140, 141], [142, 144], [144, 145], [146, 150], [150, 151], [151, 152]]}
{"doc_key": "ai-dev-190", "ner": [[7, 9, "field"], [13, 14, "field"], [19, 20, "field"], [29, 31, "field"], [35, 36, "field"], [39, 41, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [], "relations_mapping_to_source": [], "sentence": ["Elle", "fait", "appel", "aux", "domaines", "de", "la", "vision", "par", "ordinateur", "ou", "de", "la", "vision", "artificielle", ",", "et", "de", "l'", "imagerie", "m\u00e9dicale", ",", "et", "fait", "un", "usage", "intensif", "de", "la", "reconnaissance", "des", "formes", ",", "de", "la", "g\u00e9om\u00e9trie", "num\u00e9rique", "et", "du", "traitement", "du", "signal", "."], "sentence-detokenized": "Elle fait appel aux domaines de la vision par ordinateur ou de la vision artificielle, et de l'imagerie m\u00e9dicale, et fait un usage intensif de la reconnaissance des formes, de la g\u00e9om\u00e9trie num\u00e9rique et du traitement du signal.", "token2charspan": [[0, 4], [5, 9], [10, 15], [16, 19], [20, 28], [29, 31], [32, 34], [35, 41], [42, 45], [46, 56], [57, 59], [60, 62], [63, 65], [66, 72], [73, 85], [85, 86], [87, 89], [90, 92], [93, 95], [95, 103], [104, 112], [112, 113], [114, 116], [117, 121], [122, 124], [125, 130], [131, 139], [140, 142], [143, 145], [146, 160], [161, 164], [165, 171], [171, 172], [173, 175], [176, 178], [179, 188], [189, 198], [199, 201], [202, 204], [205, 215], [216, 218], [219, 225], [225, 226]]}
{"doc_key": "ai-dev-191", "ner": [[2, 5, "product"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Dans", "un", "syst\u00e8me", "de", "reconnaissance", "faciale", ",", "par", "exemple", ",", "l'", "image", "du", "visage", "d'", "une", "personne", "constitue", "l'", "entr\u00e9e", "et", "l'", "\u00e9tiquette", "de", "sortie", "est", "le", "nom", "de", "cette", "personne", "."], "sentence-detokenized": "Dans un syst\u00e8me de reconnaissance faciale, par exemple, l'image du visage d'une personne constitue l'entr\u00e9e et l'\u00e9tiquette de sortie est le nom de cette personne.", "token2charspan": [[0, 4], [5, 7], [8, 15], [16, 18], [19, 33], [34, 41], [41, 42], [43, 46], [47, 54], [54, 55], [56, 58], [58, 63], [64, 66], [67, 73], [74, 76], [76, 79], [80, 88], [89, 98], [99, 101], [101, 107], [108, 110], [111, 113], [113, 122], [123, 125], [126, 132], [133, 136], [137, 139], [140, 143], [144, 146], [147, 152], [153, 161], [161, 162]]}
{"doc_key": "ai-dev-192", "ner": [[0, 1, "organisation"], [4, 5, "product"], [9, 10, "product"], [18, 19, "product"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[0, 1, 4, 5, "artifact", "", false, false], [4, 5, 9, 10, "part-of", "", false, false], [9, 10, 4, 5, "usage", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Apple", "Inc", "a", "introduit", "Face", "ID", "sur", "le", "fleuron", "iPhone", "X", "comme", "un", "successeur", "d'", "authentification", "biom\u00e9trique", "au", "Touch", "ID", ",", "un", "syst\u00e8me", "bas\u00e9", "sur", "les", "empreintes", "digitales", "."], "sentence-detokenized": "Apple Inc a introduit Face ID sur le fleuron iPhone X comme un successeur d'authentification biom\u00e9trique au Touch ID, un syst\u00e8me bas\u00e9 sur les empreintes digitales.", "token2charspan": [[0, 5], [6, 9], [10, 11], [12, 21], [22, 26], [27, 29], [30, 33], [34, 36], [37, 44], [45, 51], [52, 53], [54, 59], [60, 62], [63, 73], [74, 76], [76, 92], [93, 104], [105, 107], [108, 113], [114, 116], [116, 117], [118, 120], [121, 128], [129, 133], [134, 137], [138, 141], [142, 152], [153, 162], [162, 163]]}
{"doc_key": "ai-dev-193", "ner": [[5, 6, "metrics"], [9, 10, "metrics"], [24, 27, "metrics"], [30, 32, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [], "relations_mapping_to_source": [], "sentence": ["Vous", "pouvez", "\u00e9galement", "combiner", "la", "mesure", "F", "avec", "le", "carr\u00e9", "R", "\u00e9valu\u00e9", "pour", "la", "sortie", "brute", "du", "mod\u00e8le", "et", "la", "cible", ",", "ou", "la", "matrice", "co\u00fbt", "/", "gain", "avec", "le", "coefficient", "de", "corr\u00e9lation", ",", "etc."], "sentence-detokenized": "Vous pouvez \u00e9galement combiner la mesure F avec le carr\u00e9 R \u00e9valu\u00e9 pour la sortie brute du mod\u00e8le et la cible, ou la matrice co\u00fbt/gain avec le coefficient de corr\u00e9lation, etc.", "token2charspan": [[0, 4], [5, 11], [12, 21], [22, 30], [31, 33], [34, 40], [41, 42], [43, 47], [48, 50], [51, 56], [57, 58], [59, 65], [66, 70], [71, 73], [74, 80], [81, 86], [87, 89], [90, 96], [97, 99], [100, 102], [103, 108], [108, 109], [110, 112], [113, 115], [116, 123], [124, 128], [128, 129], [129, 133], [134, 138], [139, 141], [142, 153], [154, 156], [157, 168], [168, 169], [170, 174]]}
{"doc_key": "ai-dev-194", "ner": [[1, 7, "conference"], [11, 13, "location"], [16, 16, "location"], [20, 24, "location"], [26, 26, "location"], [29, 29, "country"], [35, 39, "location"], [43, 48, "location"], [50, 50, "location"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "relations": [[1, 7, 11, 13, "physical", "", false, false], [1, 7, 20, 24, "physical", "", false, false], [1, 7, 35, 39, "physical", "", false, false], [1, 7, 43, 48, "physical", "", false, false], [11, 13, 16, 16, "physical", "", false, false], [20, 24, 26, 26, "physical", "", false, false], [26, 26, 29, 29, "physical", "", false, false], [35, 39, 50, 50, "physical", "", false, false], [43, 48, 50, 50, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "sentence": ["L'", "\u00e9dition", "espagnole", "de", "la", "Campus", "Party", "s'", "est", "tenue", "au", "Colegio", "Miguel", "Hern\u00e1ndez", ",", "\u00e0", "Ceulaj", "et", "\u00e0", "l'", "ar\u00e8ne", "sportive", "municipale", "de", "Benalm\u00e1dena", "\u00e0", "M\u00e1laga", ",", "en", "Espagne", ",", "ainsi", "qu'", "\u00e0", "la", "foire", "du", "comt\u00e9", "de", "Valence", "et", "\u00e0", "la", "Cit\u00e9", "des", "arts", "et", "des", "sciences", "de", "Valence", "au", "cours", "des", "15", "derni\u00e8res", "ann\u00e9es", "."], "sentence-detokenized": "L'\u00e9dition espagnole de la Campus Party s'est tenue au Colegio Miguel Hern\u00e1ndez, \u00e0 Ceulaj et \u00e0 l'ar\u00e8ne sportive municipale de Benalm\u00e1dena \u00e0 M\u00e1laga, en Espagne, ainsi qu'\u00e0 la foire du comt\u00e9 de Valence et \u00e0 la Cit\u00e9 des arts et des sciences de Valence au cours des 15 derni\u00e8res ann\u00e9es.", "token2charspan": [[0, 2], [2, 9], [10, 19], [20, 22], [23, 25], [26, 32], [33, 38], [39, 41], [41, 44], [45, 50], [51, 53], [54, 61], [62, 68], [69, 78], [78, 79], [80, 81], [82, 88], [89, 91], [92, 93], [94, 96], [96, 101], [102, 110], [111, 121], [122, 124], [125, 136], [137, 138], [139, 145], [145, 146], [147, 149], [150, 157], [157, 158], [159, 164], [165, 168], [168, 169], [170, 172], [173, 178], [179, 181], [182, 187], [188, 190], [191, 198], [199, 201], [202, 203], [204, 206], [207, 211], [212, 215], [216, 220], [221, 223], [224, 227], [228, 236], [237, 239], [240, 247], [248, 250], [251, 256], [257, 260], [261, 263], [264, 273], [274, 280], [280, 281]]}
{"doc_key": "ai-dev-195", "ner": [[0, 0, "product"], [21, 21, "programlang"], [26, 26, "product"], [28, 28, "product"], [31, 31, "programlang"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[21, 21, 0, 0, "general-affiliation", "", false, false], [26, 26, 21, 21, "part-of", "", false, false], [28, 28, 21, 21, "part-of", "", false, false], [31, 31, 0, 0, "general-affiliation", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["gnuplot", "peut", "\u00eatre", "utilis\u00e9", "\u00e0", "partir", "de", "divers", "langages", "de", "programmation", "pour", "repr\u00e9senter", "des", "donn\u00e9es", "sous", "forme", "de", "graphiques", ",", "notamment", "Perl", "(", "via", "les", "paquets", "PDL", "et", "CPAN", ")", ",", "Python", "(", "via", ")", "."], "sentence-detokenized": "gnuplot peut \u00eatre utilis\u00e9 \u00e0 partir de divers langages de programmation pour repr\u00e9senter des donn\u00e9es sous forme de graphiques, notamment Perl (via les paquets PDL et CPAN), Python (via ).", "token2charspan": [[0, 7], [8, 12], [13, 17], [18, 25], [26, 27], [28, 34], [35, 37], [38, 44], [45, 53], [54, 56], [57, 70], [71, 75], [76, 87], [88, 91], [92, 99], [100, 104], [105, 110], [111, 113], [114, 124], [124, 125], [126, 135], [136, 140], [141, 142], [142, 145], [146, 149], [150, 157], [158, 161], [162, 164], [165, 169], [169, 170], [170, 171], [172, 178], [179, 180], [180, 183], [184, 185], [185, 186]]}
{"doc_key": "ai-dev-196", "ner": [[3, 6, "product"], [22, 22, "conference"], [24, 24, "conference"], [38, 38, "conference"], [40, 40, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[22, 22, 3, 6, "topic", "", false, false], [24, 24, 3, 6, "topic", "", false, false], [38, 38, 3, 6, "topic", "", false, false], [40, 40, 3, 6, "topic", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Le", "domaine", "des", "syst\u00e8mes", "de", "dialogue", "parl\u00e9", "est", "assez", "vaste", "et", "comprend", "la", "recherche", "(", "pr\u00e9sent\u00e9e", "lors", "de", "conf\u00e9rences", "scientifiques", "telles", "que", "SIGdial", "et", "Interspeech", ")", "et", "un", "vaste", "secteur", "industriel", "(", "avec", "ses", "propres", "r\u00e9unions", "telles", "que", "SpeechTek", "et", "AVIOS", ")", "."], "sentence-detokenized": "Le domaine des syst\u00e8mes de dialogue parl\u00e9 est assez vaste et comprend la recherche (pr\u00e9sent\u00e9e lors de conf\u00e9rences scientifiques telles que SIGdial et Interspeech) et un vaste secteur industriel (avec ses propres r\u00e9unions telles que SpeechTek et AVIOS).", "token2charspan": [[0, 2], [3, 10], [11, 14], [15, 23], [24, 26], [27, 35], [36, 41], [42, 45], [46, 51], [52, 57], [58, 60], [61, 69], [70, 72], [73, 82], [83, 84], [84, 93], [94, 98], [99, 101], [102, 113], [114, 127], [128, 134], [135, 138], [139, 146], [147, 149], [150, 161], [161, 162], [163, 165], [166, 168], [169, 174], [175, 182], [183, 193], [194, 195], [195, 199], [200, 203], [204, 211], [212, 220], [221, 227], [228, 231], [232, 241], [242, 244], [245, 250], [250, 251], [251, 252]]}
{"doc_key": "ai-dev-197", "ner": [[3, 6, "field"], [10, 13, "task"], [16, 19, "task"], [22, 25, "task"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[10, 13, 3, 6, "part-of", "task_part_of_field", false, false], [16, 19, 3, 6, "part-of", "task_part_of_field", false, false], [22, 25, 3, 6, "part-of", "task_part_of_field", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Les", "d\u00e9fis", "du", "traitement", "du", "langage", "naturel", "concernent", "souvent", "la", "reconnaissance", "de", "la", "parole", ",", "la", "compr\u00e9hension", "du", "langage", "naturel", "et", "la", "g\u00e9n\u00e9ration", "du", "langage", "naturel", "."], "sentence-detokenized": "Les d\u00e9fis du traitement du langage naturel concernent souvent la reconnaissance de la parole, la compr\u00e9hension du langage naturel et la g\u00e9n\u00e9ration du langage naturel.", "token2charspan": [[0, 3], [4, 9], [10, 12], [13, 23], [24, 26], [27, 34], [35, 42], [43, 53], [54, 61], [62, 64], [65, 79], [80, 82], [83, 85], [86, 92], [92, 93], [94, 96], [97, 110], [111, 113], [114, 121], [122, 129], [130, 132], [133, 135], [136, 146], [147, 149], [150, 157], [158, 165], [165, 166]]}
{"doc_key": "ai-dev-198", "ner": [[5, 5, "product"], [7, 10, "product"], [44, 45, "task"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[5, 5, 7, 10, "part-of", "", false, false], [5, 5, 44, 45, "usage", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Ces", "syst\u00e8mes", ",", "tels", "que", "Siri", "du", "syst\u00e8me", "d'", "exploitation", "iOS", ",", "fonctionnent", "selon", "une", "technique", "de", "reconnaissance", "des", "formes", "similaire", "\u00e0", "celle", "des", "syst\u00e8mes", "bas\u00e9s", "sur", "le", "texte", ",", "mais", "dans", "le", "premier", "cas", ",", "l'", "entr\u00e9e", "de", "l'", "utilisateur", "se", "fait", "par", "reconnaissance", "vocale", "."], "sentence-detokenized": "Ces syst\u00e8mes, tels que Siri du syst\u00e8me d'exploitation iOS, fonctionnent selon une technique de reconnaissance des formes similaire \u00e0 celle des syst\u00e8mes bas\u00e9s sur le texte, mais dans le premier cas, l'entr\u00e9e de l'utilisateur se fait par reconnaissance vocale.", "token2charspan": [[0, 3], [4, 12], [12, 13], [14, 18], [19, 22], [23, 27], [28, 30], [31, 38], [39, 41], [41, 53], [54, 57], [57, 58], [59, 71], [72, 77], [78, 81], [82, 91], [92, 94], [95, 109], [110, 113], [114, 120], [121, 130], [131, 132], [133, 138], [139, 142], [143, 151], [152, 157], [158, 161], [162, 164], [165, 170], [170, 171], [172, 176], [177, 181], [182, 184], [185, 192], [193, 196], [196, 197], [198, 200], [200, 206], [207, 209], [210, 212], [212, 223], [224, 226], [227, 231], [232, 235], [236, 250], [251, 257], [257, 258]]}
{"doc_key": "ai-dev-199", "ner": [[0, 5, "algorithm"], [17, 18, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[0, 5, 17, 18, "related-to", "", true, false]], "relations_mapping_to_source": [0], "sentence": ["Des", "fonctions", "d'", "aptitude", "plus", "exotiques", "qui", "explorent", "la", "granularit\u00e9", "du", "mod\u00e8le", "comprennent", "l'", "aire", "sous", "la", "courbe", "ROC", "et", "la", "mesure", "du", "rang", "."], "sentence-detokenized": "Des fonctions d'aptitude plus exotiques qui explorent la granularit\u00e9 du mod\u00e8le comprennent l'aire sous la courbe ROC et la mesure du rang.", "token2charspan": [[0, 3], [4, 13], [14, 16], [16, 24], [25, 29], [30, 39], [40, 43], [44, 53], [54, 56], [57, 68], [69, 71], [72, 78], [79, 90], [91, 93], [93, 97], [98, 102], [103, 105], [106, 112], [113, 116], [117, 119], [120, 122], [123, 129], [130, 132], [133, 137], [137, 138]]}
{"doc_key": "ai-dev-200", "ner": [[3, 6, "product"], [10, 11, "researcher"], [16, 18, "product"], [22, 25, "organisation"], [27, 27, "organisation"], [35, 40, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[3, 6, 10, 11, "origin", "", false, false], [10, 11, 22, 25, "role", "", false, false], [16, 18, 10, 11, "origin", "", false, false], [27, 27, 22, 25, "named", "", false, false], [35, 40, 22, 25, "artifact", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Le", "terme", "\"", "Web", "s\u00e9mantique", "\"", "a", "\u00e9t\u00e9", "invent\u00e9", "par", "Tim", "Berners-Lee", ",", "l'", "inventeur", "du", "World", "Wide", "Web", "et", "directeur", "du", "World", "Wide", "Web", "Consortium", "(", "W3C", ")", ",", "qui", "supervise", "le", "d\u00e9veloppement", "des", "normes", "propos\u00e9es", "pour", "le", "Web", "s\u00e9mantique", "."], "sentence-detokenized": "Le terme \"Web s\u00e9mantique\" a \u00e9t\u00e9 invent\u00e9 par Tim Berners-Lee, l'inventeur du World Wide Web et directeur du World Wide Web Consortium (W3C), qui supervise le d\u00e9veloppement des normes propos\u00e9es pour le Web s\u00e9mantique.", "token2charspan": [[0, 2], [3, 8], [9, 10], [10, 13], [14, 24], [24, 25], [26, 27], [28, 31], [32, 39], [40, 43], [44, 47], [48, 59], [59, 60], [61, 63], [63, 72], [73, 75], [76, 81], [82, 86], [87, 90], [91, 93], [94, 103], [104, 106], [107, 112], [113, 117], [118, 121], [122, 132], [133, 134], [134, 137], [137, 138], [138, 139], [140, 143], [144, 153], [154, 156], [157, 170], [171, 174], [175, 181], [182, 191], [192, 196], [197, 199], [200, 203], [204, 214], [214, 215]]}
{"doc_key": "ai-dev-201", "ner": [[0, 2, "task"], [9, 9, "task"], [17, 20, "product"], [23, 27, "product"], [29, 29, "product"], [33, 34, "product"], [42, 43, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[0, 2, 17, 20, "opposite", "", false, false], [0, 2, 23, 27, "opposite", "", false, false], [0, 2, 33, 34, "opposite", "", false, false], [0, 2, 42, 43, "part-of", "", false, false], [9, 9, 0, 2, "named", "", false, false], [29, 29, 23, 27, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["La", "traduction", "automatique", ",", "parfois", "d\u00e9sign\u00e9e", "par", "l'", "abr\u00e9viation", "MT", "(", "\u00e0", "ne", "pas", "confondre", "avec", "la", "traduction", "assist\u00e9e", "par", "ordinateur", ",", "la", "traduction", "humaine", "assist\u00e9e", "par", "ordinateur", "(", "MAHT", ")", "ou", "la", "traduction", "interactive", ")", ",", "est", "un", "sous-domaine", "de", "la", "linguistique", "informatique", "qui", "\u00e9tudie", "l'", "utilisation", "de", "logiciels", "pour", "traduire", "un", "texte", "ou", "un", "discours", "d'", "une", "langue", "\u00e0", "une", "autre", "."], "sentence-detokenized": "La traduction automatique, parfois d\u00e9sign\u00e9e par l'abr\u00e9viation MT (\u00e0 ne pas confondre avec la traduction assist\u00e9e par ordinateur, la traduction humaine assist\u00e9e par ordinateur (MAHT) ou la traduction interactive), est un sous-domaine de la linguistique informatique qui \u00e9tudie l'utilisation de logiciels pour traduire un texte ou un discours d'une langue \u00e0 une autre.", "token2charspan": [[0, 2], [3, 13], [14, 25], [25, 26], [27, 34], [35, 43], [44, 47], [48, 50], [50, 61], [62, 64], [65, 66], [66, 67], [68, 70], [71, 74], [75, 84], [85, 89], [90, 92], [93, 103], [104, 112], [113, 116], [117, 127], [127, 128], [129, 131], [132, 142], [143, 150], [151, 159], [160, 163], [164, 174], [175, 176], [176, 180], [180, 181], [182, 184], [185, 187], [188, 198], [199, 210], [210, 211], [211, 212], [213, 216], [217, 219], [220, 232], [233, 235], [236, 238], [239, 251], [252, 264], [265, 268], [269, 275], [276, 278], [278, 289], [290, 292], [293, 302], [303, 307], [308, 316], [317, 319], [320, 325], [326, 328], [329, 331], [332, 340], [341, 343], [343, 346], [347, 353], [354, 355], [356, 359], [360, 365], [365, 366]]}
{"doc_key": "ai-dev-202", "ner": [[2, 5, "product"], [11, 11, "university"], [17, 18, "researcher"], [20, 21, "researcher"], [47, 48, "location"], [50, 50, "location"], [54, 58, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[2, 5, 17, 18, "artifact", "", false, false], [2, 5, 20, 21, "artifact", "", false, false], [17, 18, 11, 11, "physical", "", false, false], [17, 18, 11, 11, "role", "", false, false], [20, 21, 11, 11, "physical", "", false, false], [20, 21, 11, 11, "role", "", false, false], [47, 48, 50, 50, "physical", "", false, false], [54, 58, 47, 48, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "sentence": ["Les", "premiers", "syst\u00e8mes", "de", "TA", "interlingue", "ont", "\u00e9galement", "\u00e9t\u00e9", "construits", "\u00e0", "Stanford", "dans", "les", "ann\u00e9es", "1970", "par", "Roger", "Schank", "et", "Yorick", "Wilks", ";", "le", "premier", "est", "devenu", "la", "base", "d'", "un", "syst\u00e8me", "commercial", "pour", "le", "transfert", "de", "fonds", ",", "et", "le", "code", "du", "second", "est", "conserv\u00e9", "au", "Computer", "Museum", "de", "Boston", "comme", "le", "premier", "syst\u00e8me", "de", "traduction", "automatique", "interlingue", "."], "sentence-detokenized": "Les premiers syst\u00e8mes de TA interlingue ont \u00e9galement \u00e9t\u00e9 construits \u00e0 Stanford dans les ann\u00e9es 1970 par Roger Schank et Yorick Wilks ; le premier est devenu la base d'un syst\u00e8me commercial pour le transfert de fonds, et le code du second est conserv\u00e9 au Computer Museum de Boston comme le premier syst\u00e8me de traduction automatique interlingue.", "token2charspan": [[0, 3], [4, 12], [13, 21], [22, 24], [25, 27], [28, 39], [40, 43], [44, 53], [54, 57], [58, 68], [69, 70], [71, 79], [80, 84], [85, 88], [89, 95], [96, 100], [101, 104], [105, 110], [111, 117], [118, 120], [121, 127], [128, 133], [134, 135], [136, 138], [139, 146], [147, 150], [151, 157], [158, 160], [161, 165], [166, 168], [168, 170], [171, 178], [179, 189], [190, 194], [195, 197], [198, 207], [208, 210], [211, 216], [216, 217], [218, 220], [221, 223], [224, 228], [229, 231], [232, 238], [239, 242], [243, 251], [252, 254], [255, 263], [264, 270], [271, 273], [274, 280], [281, 286], [287, 289], [290, 297], [298, 305], [306, 308], [309, 319], [320, 331], [332, 343], [343, 344]]}
{"doc_key": "ai-dev-203", "ner": [[0, 1, "researcher"], [9, 15, "conference"], [17, 18, "conference"], [25, 31, "conference"], [33, 34, "conference"], [39, 46, "organisation"], [58, 58, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[0, 1, 9, 15, "role", "", false, false], [0, 1, 25, 31, "role", "", false, false], [0, 1, 39, 46, "role", "", false, false], [0, 1, 58, 58, "role", "", false, false], [17, 18, 9, 15, "named", "", false, false], [33, 34, 25, 31, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["M.", "Sycara", "a", "\u00e9t\u00e9", "pr\u00e9sident", "du", "programme", "de", "la", "deuxi\u00e8me", "conf\u00e9rence", "internationale", "sur", "le", "Web", "s\u00e9mantique", "(", "ISWC", "2003", ")", ",", "pr\u00e9sident", "g\u00e9n\u00e9ral", "de", "la", "deuxi\u00e8me", "conf\u00e9rence", "internationale", "sur", "les", "agents", "autonomes", "(", "Agents", "98", ")", ",", "pr\u00e9sident", "du", "comit\u00e9", "directeur", "de", "la", "conf\u00e9rence", "sur", "les", "agents", "(", "1999-2001", ")", "et", "pr\u00e9sident", "du", "comit\u00e9", "des", "bourses", "de", "l'", "AAAI", "(", "1993-1999", ")", ";"], "sentence-detokenized": "M. Sycara a \u00e9t\u00e9 pr\u00e9sident du programme de la deuxi\u00e8me conf\u00e9rence internationale sur le Web s\u00e9mantique (ISWC 2003), pr\u00e9sident g\u00e9n\u00e9ral de la deuxi\u00e8me conf\u00e9rence internationale sur les agents autonomes (Agents 98), pr\u00e9sident du comit\u00e9 directeur de la conf\u00e9rence sur les agents (1999-2001) et pr\u00e9sident du comit\u00e9 des bourses de l'AAAI (1993-1999) ;", "token2charspan": [[0, 2], [3, 9], [10, 11], [12, 15], [16, 25], [26, 28], [29, 38], [39, 41], [42, 44], [45, 53], [54, 64], [65, 79], [80, 83], [84, 86], [87, 90], [91, 101], [102, 103], [103, 107], [108, 112], [112, 113], [113, 114], [115, 124], [125, 132], [133, 135], [136, 138], [139, 147], [148, 158], [159, 173], [174, 177], [178, 181], [182, 188], [189, 198], [199, 200], [200, 206], [207, 209], [209, 210], [210, 211], [212, 221], [222, 224], [225, 231], [232, 241], [242, 244], [245, 247], [248, 258], [259, 262], [263, 266], [267, 273], [274, 275], [275, 284], [284, 285], [286, 288], [289, 298], [299, 301], [302, 308], [309, 312], [313, 320], [321, 323], [324, 326], [326, 330], [331, 332], [332, 341], [341, 342], [343, 344]]}
{"doc_key": "ai-dev-204", "ner": [[15, 15, "conference"], [17, 20, "conference"], [10, 13, "misc"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[17, 20, 15, 15, "named", "", false, false], [10, 13, 15, 15, "part-of", "", false, false], [10, 13, 15, 15, "temporal", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["En", "2016", ",", "elle", "a", "\u00e9t\u00e9", "s\u00e9lectionn\u00e9e", "comme", "laur\u00e9ate", "du", "Lifetime", "Achievement", "Award", "de", "l'", "ACL", "(", "Association", "for", "Computational", "Linguistics", ")", "."], "sentence-detokenized": "En 2016, elle a \u00e9t\u00e9 s\u00e9lectionn\u00e9e comme laur\u00e9ate du Lifetime Achievement Award de l'ACL (Association for Computational Linguistics).", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 13], [14, 15], [16, 19], [20, 32], [33, 38], [39, 47], [48, 50], [51, 59], [60, 71], [72, 77], [78, 80], [81, 83], [83, 86], [87, 88], [88, 99], [100, 103], [104, 117], [118, 129], [129, 130], [130, 131]]}
{"doc_key": "ai-dev-205", "ner": [[0, 1, "researcher"], [3, 4, "researcher"], [6, 7, "researcher"], [10, 11, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [], "relations_mapping_to_source": [], "sentence": ["Sepp", "Hochreiter", ",", "Y.", "Bengio", ",", "P.", "Frasconi", ",", "et", "J\u00fcrgen", "Schmidhuber", "."], "sentence-detokenized": "Sepp Hochreiter, Y. Bengio, P. Frasconi, et J\u00fcrgen Schmidhuber.", "token2charspan": [[0, 4], [5, 15], [15, 16], [17, 19], [20, 26], [26, 27], [28, 30], [31, 39], [39, 40], [41, 43], [44, 50], [51, 62], [62, 63]]}
{"doc_key": "ai-dev-206", "ner": [[3, 3, "product"], [6, 8, "misc"], [10, 10, "programlang"], [19, 21, "product"], [35, 35, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[3, 3, 10, 10, "usage", "", false, false], [10, 10, 6, 8, "type-of", "", false, false], [10, 10, 19, 21, "related-to", "", false, false], [35, 35, 3, 3, "origin", "", false, true]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Par", "exemple", ",", "A.L.I.C.E.", "utilise", "un", "langage", "de", "balisage", "appel\u00e9", "AIML", ",", "qui", "est", "sp\u00e9cifique", "\u00e0", "sa", "fonction", "de", "syst\u00e8me", "de", "dialogue", ",", "et", "qui", "a", "depuis", "\u00e9t\u00e9", "adopt\u00e9", "par", "divers", "autres", "d\u00e9veloppeurs", "de", "\"", "Alicebots", "\"", "."], "sentence-detokenized": "Par exemple, A.L.I.C.E. utilise un langage de balisage appel\u00e9 AIML, qui est sp\u00e9cifique \u00e0 sa fonction de syst\u00e8me de dialogue, et qui a depuis \u00e9t\u00e9 adopt\u00e9 par divers autres d\u00e9veloppeurs de \"Alicebots\".", "token2charspan": [[0, 3], [4, 11], [11, 12], [13, 23], [24, 31], [32, 34], [35, 42], [43, 45], [46, 54], [55, 61], [62, 66], [66, 67], [68, 71], [72, 75], [76, 86], [87, 88], [89, 91], [92, 100], [101, 103], [104, 111], [112, 114], [115, 123], [123, 124], [125, 127], [128, 131], [132, 133], [134, 140], [141, 144], [145, 151], [152, 155], [156, 162], [163, 169], [170, 182], [183, 185], [186, 187], [187, 196], [196, 197], [197, 198]]}
{"doc_key": "ai-dev-207", "ner": [[10, 16, "conference"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["En", "2000", ",", "elle", "a", "\u00e9t\u00e9", "\u00e9lue", "membre", "de", "l'", "Association", "for", "the", "Advancement", "of", "Artificial", "Intelligence", "."], "sentence-detokenized": "En 2000, elle a \u00e9t\u00e9 \u00e9lue membre de l'Association for the Advancement of Artificial Intelligence.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 13], [14, 15], [16, 19], [20, 24], [25, 31], [32, 34], [35, 37], [37, 48], [49, 52], [53, 56], [57, 68], [69, 71], [72, 82], [83, 95], [95, 96]]}
{"doc_key": "ai-dev-208", "ner": [[0, 5, "misc"], [7, 7, "misc"], [13, 20, "misc"], [30, 31, "algorithm"], [42, 43, "field"], [47, 49, "field"], [53, 55, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[0, 5, 13, 20, "type-of", "", false, false], [0, 5, 42, 43, "related-to", "performs", true, false], [0, 5, 47, 49, "related-to", "performs", true, false], [0, 5, 53, 55, "related-to", "performs", true, false], [7, 7, 0, 5, "named", "", false, false], [30, 31, 13, 20, "part-of", "", true, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["Les", "syst\u00e8mes", "de", "classification", "par", "apprentissage", "(", "LCS", ")", "sont", "une", "famille", "d'", "algorithmes", "d'", "apprentissage", "automatique", "bas\u00e9s", "sur", "des", "r\u00e8gles", "qui", "combinent", "un", "composant", "de", "d\u00e9couverte", ",", "g\u00e9n\u00e9ralement", "un", "algorithme", "g\u00e9n\u00e9tique", ",", "avec", "un", "composant", "d'", "apprentissage", ",", "r\u00e9alisant", "soit", "un", "apprentissage", "supervis\u00e9", ",", "soit", "un", "apprentissage", "par", "renforcement", ",", "soit", "un", "apprentissage", "non", "supervis\u00e9", "."], "sentence-detokenized": "Les syst\u00e8mes de classification par apprentissage (LCS) sont une famille d'algorithmes d'apprentissage automatique bas\u00e9s sur des r\u00e8gles qui combinent un composant de d\u00e9couverte, g\u00e9n\u00e9ralement un algorithme g\u00e9n\u00e9tique, avec un composant d'apprentissage, r\u00e9alisant soit un apprentissage supervis\u00e9, soit un apprentissage par renforcement, soit un apprentissage non supervis\u00e9.", "token2charspan": [[0, 3], [4, 12], [13, 15], [16, 30], [31, 34], [35, 48], [49, 50], [50, 53], [53, 54], [55, 59], [60, 63], [64, 71], [72, 74], [74, 85], [86, 88], [88, 101], [102, 113], [114, 119], [120, 123], [124, 127], [128, 134], [135, 138], [139, 148], [149, 151], [152, 161], [162, 164], [165, 175], [175, 176], [177, 189], [190, 192], [193, 203], [204, 213], [213, 214], [215, 219], [220, 222], [223, 232], [233, 235], [235, 248], [248, 249], [250, 259], [260, 264], [265, 267], [268, 281], [282, 291], [291, 292], [293, 297], [298, 300], [301, 314], [315, 318], [319, 331], [331, 332], [333, 337], [338, 340], [341, 354], [355, 358], [359, 368], [368, 369]]}
{"doc_key": "ai-dev-209", "ner": [[17, 19, "algorithm"], [21, 21, "algorithm"], [29, 31, "algorithm"], [34, 34, "misc"], [45, 49, "algorithm"], [57, 62, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[17, 19, 29, 31, "origin", "", false, false], [17, 19, 34, 34, "usage", "", false, false], [21, 21, 17, 19, "named", "", false, false], [45, 49, 34, 34, "type-of", "", false, false], [45, 49, 57, 62, "compare", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Les", "param\u00e8tres", "inconnus", "de", "chaque", "vecteur", "\u03b2subk", "/", "sub", "sont", "g\u00e9n\u00e9ralement", "estim\u00e9s", "conjointement", "par", "une", "estimation", "par", "maximum", "a", "posteriori", "(", "MAP", ")", ",", "qui", "est", "une", "extension", "du", "maximum", "de", "vraisemblance", "utilisant", "une", "r\u00e9gularisation", "des", "poids", "pour", "\u00e9viter", "les", "solutions", "pathologiques", "(", "g\u00e9n\u00e9ralement", "une", "fonction", "de", "r\u00e9gularisation", "au", "carr\u00e9", ",", "ce", "qui", "\u00e9quivaut", "\u00e0", "placer", "une", "distribution", "ant\u00e9rieure", "gaussienne", "\u00e0", "moyenne", "nulle", "sur", "les", "poids", ",", "mais", "d'", "autres", "distributions", "sont", "\u00e9galement", "possibles", ")", "."], "sentence-detokenized": "Les param\u00e8tres inconnus de chaque vecteur \u03b2subk / sub sont g\u00e9n\u00e9ralement estim\u00e9s conjointement par une estimation par maximum a posteriori (MAP), qui est une extension du maximum de vraisemblance utilisant une r\u00e9gularisation des poids pour \u00e9viter les solutions pathologiques (g\u00e9n\u00e9ralement une fonction de r\u00e9gularisation au carr\u00e9, ce qui \u00e9quivaut \u00e0 placer une distribution ant\u00e9rieure gaussienne \u00e0 moyenne nulle sur les poids, mais d'autres distributions sont \u00e9galement possibles).", "token2charspan": [[0, 3], [4, 14], [15, 23], [24, 26], [27, 33], [34, 41], [42, 47], [48, 49], [50, 53], [54, 58], [59, 71], [72, 79], [80, 93], [94, 97], [98, 101], [102, 112], [113, 116], [117, 124], [125, 126], [127, 137], [138, 139], [139, 142], [142, 143], [143, 144], [145, 148], [149, 152], [153, 156], [157, 166], [167, 169], [170, 177], [178, 180], [181, 194], [195, 204], [205, 208], [209, 223], [224, 227], [228, 233], [234, 238], [239, 245], [246, 249], [250, 259], [260, 273], [274, 275], [275, 287], [288, 291], [292, 300], [301, 303], [304, 318], [319, 321], [322, 327], [327, 328], [329, 331], [332, 335], [336, 344], [345, 346], [347, 353], [354, 357], [358, 370], [371, 381], [382, 392], [393, 394], [395, 402], [403, 408], [409, 412], [413, 416], [417, 422], [422, 423], [424, 428], [429, 431], [431, 437], [438, 451], [452, 456], [457, 466], [467, 476], [476, 477], [477, 478]]}
{"doc_key": "ai-dev-210", "ner": [[13, 14, "researcher"], [11, 11, "product"]], "ner_mapping_to_source": [0, 1], "relations": [[11, 11, 13, 14, "artifact", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["La", "structure", "hi\u00e9rarchique", "des", "mots", "a", "\u00e9t\u00e9", "explicitement", "cartographi\u00e9e", "dans", "le", "Wordnet", "de", "George", "Miller", "."], "sentence-detokenized": "La structure hi\u00e9rarchique des mots a \u00e9t\u00e9 explicitement cartographi\u00e9e dans le Wordnet de George Miller.", "token2charspan": [[0, 2], [3, 12], [13, 25], [26, 29], [30, 34], [35, 36], [37, 40], [41, 54], [55, 68], [69, 73], [74, 76], [77, 84], [85, 87], [88, 94], [95, 101], [101, 102]]}
{"doc_key": "ai-dev-211", "ner": [[9, 16, "conference"], [27, 32, "task"]], "ner_mapping_to_source": [0, 1], "relations": [[27, 32, 9, 16, "topic", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Une", "illustration", "de", "leurs", "capacit\u00e9s", "est", "donn\u00e9e", "par", "le", "d\u00e9fi", "de", "reconnaissance", "visuelle", "\u00e0", "grande", "\u00e9chelle", "ImageNet", ";", "il", "s'", "agit", "d'", "une", "r\u00e9f\u00e9rence", "en", "mati\u00e8re", "de", "classification", "et", "de", "d\u00e9tection", "d'", "objets", ",", "avec", "des", "millions", "d'", "images", "et", "des", "centaines", "de", "classes", "d'", "objets", "."], "sentence-detokenized": "Une illustration de leurs capacit\u00e9s est donn\u00e9e par le d\u00e9fi de reconnaissance visuelle \u00e0 grande \u00e9chelle ImageNet ; il s'agit d'une r\u00e9f\u00e9rence en mati\u00e8re de classification et de d\u00e9tection d'objets, avec des millions d'images et des centaines de classes d'objets.", "token2charspan": [[0, 3], [4, 16], [17, 19], [20, 25], [26, 35], [36, 39], [40, 46], [47, 50], [51, 53], [54, 58], [59, 61], [62, 76], [77, 85], [86, 87], [88, 94], [95, 102], [103, 111], [112, 113], [114, 116], [117, 119], [119, 123], [124, 126], [126, 129], [130, 139], [140, 142], [143, 150], [151, 153], [154, 168], [169, 171], [172, 174], [175, 184], [185, 187], [187, 193], [193, 194], [195, 199], [200, 203], [204, 212], [213, 215], [215, 221], [222, 224], [225, 228], [229, 238], [239, 241], [242, 249], [250, 252], [252, 258], [258, 259]]}
{"doc_key": "ai-dev-212", "ner": [[2, 4, "misc"], [26, 26, "misc"], [35, 37, "person"], [30, 30, "misc"], [47, 49, "person"], [41, 42, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[26, 26, 2, 4, "general-affiliation", "", false, false], [30, 30, 2, 4, "general-affiliation", "", false, false], [30, 30, 35, 37, "artifact", "", false, false], [41, 42, 2, 4, "general-affiliation", "", false, false], [41, 42, 47, 49, "artifact", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Dans", "la", "science-fiction", ",", "les", "robots", "\u00e0", "l'", "apparence", "f\u00e9minine", "sont", "souvent", "produits", "pour", "servir", "de", "domestiques", "et", "d'", "esclaves", "sexuels", ",", "comme", "dans", "le", "film", "Westworld", ",", "le", "roman", "Fairyland", "(", "1995", ")", "de", "Paul", "J.", "McAuley", "et", "la", "nouvelle", "Helen", "O'Loy", "(", "1938", ")", "de", "Lester", "del", "Rey", ",", "et", "parfois", "comme", "guerriers", ",", "tueurs", "ou", "ouvriers", "."], "sentence-detokenized": "Dans la science-fiction, les robots \u00e0 l'apparence f\u00e9minine sont souvent produits pour servir de domestiques et d'esclaves sexuels, comme dans le film Westworld, le roman Fairyland (1995) de Paul J. McAuley et la nouvelle Helen O'Loy (1938) de Lester del Rey, et parfois comme guerriers, tueurs ou ouvriers.", "token2charspan": [[0, 4], [5, 7], [8, 23], [23, 24], [25, 28], [29, 35], [36, 37], [38, 40], [40, 49], [50, 58], [59, 63], [64, 71], [72, 80], [81, 85], [86, 92], [93, 95], [96, 107], [108, 110], [111, 113], [113, 121], [122, 129], [129, 130], [131, 136], [137, 141], [142, 144], [145, 149], [150, 159], [159, 160], [161, 163], [164, 169], [170, 179], [180, 181], [181, 185], [185, 186], [187, 189], [190, 194], [195, 197], [198, 205], [206, 208], [209, 211], [212, 220], [221, 226], [227, 232], [233, 234], [234, 238], [238, 239], [240, 242], [243, 249], [250, 253], [254, 257], [257, 258], [259, 261], [262, 269], [270, 275], [276, 285], [285, 286], [287, 293], [294, 296], [297, 305], [305, 306]]}
{"doc_key": "ai-dev-213", "ner": [[0, 3, "task"], [6, 7, "task"], [10, 11, "task"]], "ner_mapping_to_source": [0, 1, 2], "relations": [], "relations_mapping_to_source": [], "sentence": ["la", "r\u00e9ponse", "aux", "questions", ",", "la", "reconnaissance", "vocale", "et", "la", "traduction", "automatique", "."], "sentence-detokenized": "la r\u00e9ponse aux questions, la reconnaissance vocale et la traduction automatique.", "token2charspan": [[0, 2], [3, 10], [11, 14], [15, 24], [24, 25], [26, 28], [29, 43], [44, 50], [51, 53], [54, 56], [57, 67], [68, 79], [79, 80]]}
{"doc_key": "ai-dev-214", "ner": [[5, 6, "researcher"], [9, 18, "organisation"], [21, 24, "location"], [27, 27, "location"], [31, 31, "location"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[5, 6, 9, 18, "role", "", false, false], [9, 18, 21, 24, "physical", "", false, false], [21, 24, 27, 27, "physical", "", false, false], [27, 27, 31, 31, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Dans", "son", "article", "fondateur", ",", "Harry", "Blum", ",", "des", "laboratoires", "de", "recherche", "Cambridge", "de", "l'", "arm\u00e9e", "de", "l'", "air", "\u00e0", "la", "base", "a\u00e9rienne", "de", "Hanscom", ",", "\u00e0", "Bedford", ",", "dans", "le", "Massachusetts", ",", "a", "d\u00e9fini", "un", "axe", "m\u00e9dian", "pour", "calculer", "le", "squelette", "d'", "une", "forme", ",", "en", "utilisant", "un", "mod\u00e8le", "intuitif", "de", "propagation", "du", "feu", "sur", "un", "champ", "d'", "herbe", ",", "o\u00f9", "le", "champ", "a", "la", "forme", "de", "la", "forme", "donn\u00e9e", "."], "sentence-detokenized": "Dans son article fondateur, Harry Blum, des laboratoires de recherche Cambridge de l'arm\u00e9e de l'air \u00e0 la base a\u00e9rienne de Hanscom, \u00e0 Bedford, dans le Massachusetts, a d\u00e9fini un axe m\u00e9dian pour calculer le squelette d'une forme, en utilisant un mod\u00e8le intuitif de propagation du feu sur un champ d'herbe, o\u00f9 le champ a la forme de la forme donn\u00e9e.", "token2charspan": [[0, 4], [5, 8], [9, 16], [17, 26], [26, 27], [28, 33], [34, 38], [38, 39], [40, 43], [44, 56], [57, 59], [60, 69], [70, 79], [80, 82], [83, 85], [85, 90], [91, 93], [94, 96], [96, 99], [100, 101], [102, 104], [105, 109], [110, 118], [119, 121], [122, 129], [129, 130], [131, 132], [133, 140], [140, 141], [142, 146], [147, 149], [150, 163], [163, 164], [165, 166], [167, 173], [174, 176], [177, 180], [181, 187], [188, 192], [193, 201], [202, 204], [205, 214], [215, 217], [217, 220], [221, 226], [226, 227], [228, 230], [231, 240], [241, 243], [244, 250], [251, 259], [260, 262], [263, 274], [275, 277], [278, 281], [282, 285], [286, 288], [289, 294], [295, 297], [297, 302], [302, 303], [304, 306], [307, 309], [310, 315], [316, 317], [318, 320], [321, 326], [327, 329], [330, 332], [333, 338], [339, 345], [345, 346]]}
{"doc_key": "ai-dev-215", "ner": [[18, 18, "algorithm"], [20, 20, "algorithm"], [23, 23, "algorithm"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[18, 18, 23, 23, "compare", "", false, false], [20, 20, 23, 23, "compare", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Cependant", ",", "contrairement", "aux", "algorithmes", "de", "boosting", "qui", "minimisent", "analytiquement", "une", "fonction", "de", "perte", "convexe", "(", "par", "exemple", "AdaBoost", "et", "LogitBoost", ")", ",", "BrownBoost", "r\u00e9sout", "un", "syst\u00e8me", "de", "deux", "\u00e9quations", "et", "deux", "inconnues", "en", "utilisant", "des", "m\u00e9thodes", "num\u00e9riques", "standard", "."], "sentence-detokenized": "Cependant, contrairement aux algorithmes de boosting qui minimisent analytiquement une fonction de perte convexe (par exemple AdaBoost et LogitBoost), BrownBoost r\u00e9sout un syst\u00e8me de deux \u00e9quations et deux inconnues en utilisant des m\u00e9thodes num\u00e9riques standard.", "token2charspan": [[0, 9], [9, 10], [11, 24], [25, 28], [29, 40], [41, 43], [44, 52], [53, 56], [57, 67], [68, 82], [83, 86], [87, 95], [96, 98], [99, 104], [105, 112], [113, 114], [114, 117], [118, 125], [126, 134], [135, 137], [138, 148], [148, 149], [149, 150], [151, 161], [162, 168], [169, 171], [172, 179], [180, 182], [183, 187], [188, 197], [198, 200], [201, 205], [206, 215], [216, 218], [219, 228], [229, 232], [233, 241], [242, 252], [253, 261], [261, 262]]}
{"doc_key": "ai-dev-216", "ner": [[0, 0, "researcher"], [11, 16, "misc"], [22, 28, "conference"], [30, 30, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[0, 0, 11, 16, "win-defeat", "", false, false], [0, 0, 22, 28, "role", "", false, false], [30, 30, 22, 28, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Getoor", "a", "re\u00e7u", "plusieurs", "prix", "pour", "les", "meilleurs", "articles", ",", "une", "bourse", "de", "carri\u00e8re", "de", "la", "NSF", "et", "est", "membre", "de", "l'", "Association", "for", "the", "Advancement", "of", "Artificial", "Intelligence", "(", "AAAI", ")", "."], "sentence-detokenized": "Getoor a re\u00e7u plusieurs prix pour les meilleurs articles, une bourse de carri\u00e8re de la NSF et est membre de l'Association for the Advancement of Artificial Intelligence (AAAI).", "token2charspan": [[0, 6], [7, 8], [9, 13], [14, 23], [24, 28], [29, 33], [34, 37], [38, 47], [48, 56], [56, 57], [58, 61], [62, 68], [69, 71], [72, 80], [81, 83], [84, 86], [87, 90], [91, 93], [94, 97], [98, 104], [105, 107], [108, 110], [110, 121], [122, 125], [126, 129], [130, 141], [142, 144], [145, 155], [156, 168], [169, 170], [170, 174], [174, 175], [175, 176]]}
{"doc_key": "ai-dev-217", "ner": [[0, 1, "misc"], [6, 10, "misc"], [15, 16, "misc"], [21, 25, "misc"], [30, 31, "misc"], [35, 39, "university"], [44, 52, "misc"], [57, 66, "misc"], [71, 75, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "relations": [], "relations_mapping_to_source": [], "sentence": ["ACM", "Fellow", "(", "2015", ")", "br", "Association", "for", "Computational", "Linguistics", "Fellow", "(", "2011", ")", "br", "AAAI", "Fellow", "(", "1994", ")", "br", "International", "Speech", "Communication", "Association", "Fellow", "(", "2011", ")", "br", "Doctorat", "honorifique", "(", "Hedersdoktor", ")", "KTH", "Royal", "Institute", "of", "Technology", "(", "2007", ")", "br", "Columbia", "Engineering", "School", "Alumni", "Association", "Distinguished", "Faculty", "Teaching", "award", "(", "2009", ")", "br", "IEEE", "James", "L", ".", "Flanagan", "Speech", "and", "Audio", "Processing", "Award", "(", "2011", ")", "br", "ISCA", "Medal", "for", "Scientific", "Achievement", "(", "2011", ")"], "sentence-detokenized": "ACM Fellow (2015) br Association for Computational Linguistics Fellow (2011) br AAAI Fellow (1994) br International Speech Communication Association Fellow (2011) br Doctorat honorifique (Hedersdoktor) KTH Royal Institute of Technology (2007) br Columbia Engineering School Alumni Association Distinguished Faculty Teaching award (2009) br IEEE James L. Flanagan Speech and Audio Processing Award (2011) br ISCA Medal for Scientific Achievement (2011)", "token2charspan": [[0, 3], [4, 10], [11, 12], [12, 16], [16, 17], [18, 20], [21, 32], [33, 36], [37, 50], [51, 62], [63, 69], [70, 71], [71, 75], [75, 76], [77, 79], [80, 84], [85, 91], [92, 93], [93, 97], [97, 98], [99, 101], [102, 115], [116, 122], [123, 136], [137, 148], [149, 155], [156, 157], [157, 161], [161, 162], [163, 165], [166, 174], [175, 186], [187, 188], [188, 200], [200, 201], [202, 205], [206, 211], [212, 221], [222, 224], [225, 235], [236, 237], [237, 241], [241, 242], [243, 245], [246, 254], [255, 266], [267, 273], [274, 280], [281, 292], [293, 306], [307, 314], [315, 323], [324, 329], [330, 331], [331, 335], [335, 336], [337, 339], [340, 344], [345, 350], [351, 352], [352, 353], [354, 362], [363, 369], [370, 373], [374, 379], [380, 390], [391, 396], [397, 398], [398, 402], [402, 403], [404, 406], [407, 411], [412, 417], [418, 421], [422, 432], [433, 444], [445, 446], [446, 450], [450, 451]]}
{"doc_key": "ai-dev-218", "ner": [[8, 8, "university"], [18, 22, "task"], [35, 39, "metrics"], [50, 53, "task"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[35, 39, 50, 53, "opposite", "", true, false]], "relations_mapping_to_source": [0], "sentence": ["Un", "r\u00e9sultat", "frustrant", "de", "la", "m\u00eame", "\u00e9tude", "de", "Stanford", "(", "et", "d'", "autres", "tentatives", "d'", "am\u00e9lioration", "de", "la", "traduction", "par", "reconnaissance", "de", "noms", ")", "est", "que", ",", "bien", "souvent", ",", "une", "diminution", "des", "scores", "de", "doublure", "de", "l'", "\u00e9valuation", "bilingue", "pour", "la", "traduction", "r\u00e9sulte", "de", "l'", "inclusion", "de", "m\u00e9thodes", "de", "traduction", "d'", "entit\u00e9s", "nomm\u00e9es", "."], "sentence-detokenized": "Un r\u00e9sultat frustrant de la m\u00eame \u00e9tude de Stanford (et d'autres tentatives d'am\u00e9lioration de la traduction par reconnaissance de noms) est que, bien souvent, une diminution des scores de doublure de l'\u00e9valuation bilingue pour la traduction r\u00e9sulte de l'inclusion de m\u00e9thodes de traduction d'entit\u00e9s nomm\u00e9es.", "token2charspan": [[0, 2], [3, 11], [12, 21], [22, 24], [25, 27], [28, 32], [33, 38], [39, 41], [42, 50], [51, 52], [52, 54], [55, 57], [57, 63], [64, 74], [75, 77], [77, 89], [90, 92], [93, 95], [96, 106], [107, 110], [111, 125], [126, 128], [129, 133], [133, 134], [135, 138], [139, 142], [142, 143], [144, 148], [149, 156], [156, 157], [158, 161], [162, 172], [173, 176], [177, 183], [184, 186], [187, 195], [196, 198], [199, 201], [201, 211], [212, 220], [221, 225], [226, 228], [229, 239], [240, 247], [248, 250], [251, 253], [253, 262], [263, 265], [266, 274], [275, 277], [278, 288], [289, 291], [291, 298], [299, 306], [306, 307]]}
{"doc_key": "ai-dev-219", "ner": [[0, 0, "organisation"], [13, 15, "organisation"], [19, 28, "university"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[0, 0, 13, 15, "role", "works_with", false, false], [0, 0, 19, 28, "role", "works_with", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Medtronic", "utilise", "les", "donn\u00e9es", "PM", "collect\u00e9es", "et", "collabore", "avec", "des", "chercheurs", "de", "l'", "h\u00f4pital", "Johns", "Hopkins", "et", "de", "l'", "\u00e9cole", "de", "m\u00e9decine", "de", "l'", "universit\u00e9", "de", "Washington", "afin", "d'", "aider", "\u00e0", "r\u00e9pondre", "\u00e0", "des", "questions", "sp\u00e9cifiques", "sur", "les", "maladies", "cardiaques", ",", "par", "exemple", "pour", "savoir", "si", "un", "c\u0153ur", "faible", "provoque", "des", "arythmies", "ou", "vice", "versa", "."], "sentence-detokenized": "Medtronic utilise les donn\u00e9es PM collect\u00e9es et collabore avec des chercheurs de l'h\u00f4pital Johns Hopkins et de l'\u00e9cole de m\u00e9decine de l'universit\u00e9 de Washington afin d'aider \u00e0 r\u00e9pondre \u00e0 des questions sp\u00e9cifiques sur les maladies cardiaques, par exemple pour savoir si un c\u0153ur faible provoque des arythmies ou vice versa.", "token2charspan": [[0, 9], [10, 17], [18, 21], [22, 29], [30, 32], [33, 43], [44, 46], [47, 56], [57, 61], [62, 65], [66, 76], [77, 79], [80, 82], [82, 89], [90, 95], [96, 103], [104, 106], [107, 109], [110, 112], [112, 117], [118, 120], [121, 129], [130, 132], [133, 135], [135, 145], [146, 148], [149, 159], [160, 164], [165, 167], [167, 172], [173, 174], [175, 183], [184, 185], [186, 189], [190, 199], [200, 211], [212, 215], [216, 219], [220, 228], [229, 239], [239, 240], [241, 244], [245, 252], [253, 257], [258, 264], [265, 267], [268, 270], [271, 275], [276, 282], [283, 291], [292, 295], [296, 305], [306, 308], [309, 313], [314, 319], [319, 320]]}
{"doc_key": "ai-dev-220", "ner": [[6, 6, "organisation"], [8, 8, "misc"], [11, 12, "person"], [14, 15, "person"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[8, 8, 6, 6, "artifact", "made_by_studio", false, false], [11, 12, 8, 8, "role", "", false, false], [14, 15, 8, 8, "role", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Ensuite", ",", "le", "premier", "film", "de", "Paramount", ",", "Sangaree", ",", "avec", "Fernando", "Lamas", "et", "Arlene", "Dahl", "."], "sentence-detokenized": "Ensuite, le premier film de Paramount, Sangaree, avec Fernando Lamas et Arlene Dahl.", "token2charspan": [[0, 7], [7, 8], [9, 11], [12, 19], [20, 24], [25, 27], [28, 37], [37, 38], [39, 47], [47, 48], [49, 53], [54, 62], [63, 68], [69, 71], [72, 78], [79, 83], [83, 84]]}
{"doc_key": "ai-dev-221", "ner": [[0, 0, "programlang"], [11, 13, "researcher"], [15, 16, "researcher"], [22, 23, "organisation"], [27, 29, "university"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[0, 0, 11, 13, "origin", "", false, false], [0, 0, 15, 16, "origin", "", false, false], [11, 13, 22, 23, "physical", "", false, false], [11, 13, 22, 23, "role", "", false, false], [15, 16, 27, 29, "physical", "", false, false], [15, 16, 27, 29, "role", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["KRL", "est", "un", "langage", "de", "repr\u00e9sentation", "des", "connaissances", ",", "d\u00e9velopp\u00e9", "par", "Daniel", "G.", "Bobrow", "et", "Terry", "Winograd", "lorsqu'", "ils", "travaillaient", "respectivement", "au", "Xerox", "PARC", "et", "\u00e0", "l'", "Universit\u00e9", "de", "Stanford", "."], "sentence-detokenized": "KRL est un langage de repr\u00e9sentation des connaissances, d\u00e9velopp\u00e9 par Daniel G. Bobrow et Terry Winograd lorsqu'ils travaillaient respectivement au Xerox PARC et \u00e0 l'Universit\u00e9 de Stanford.", "token2charspan": [[0, 3], [4, 7], [8, 10], [11, 18], [19, 21], [22, 36], [37, 40], [41, 54], [54, 55], [56, 65], [66, 69], [70, 76], [77, 79], [80, 86], [87, 89], [90, 95], [96, 104], [105, 112], [112, 115], [116, 129], [130, 144], [145, 147], [148, 153], [154, 158], [159, 161], [162, 163], [164, 166], [166, 176], [177, 179], [180, 188], [188, 189]]}
{"doc_key": "ai-dev-222", "ner": [[3, 15, "conference"], [19, 20, "researcher"], [22, 23, "researcher"], [25, 26, "researcher"], [28, 30, "researcher"], [39, 40, "task"], [45, 48, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[3, 15, 39, 40, "topic", "", true, false], [19, 20, 3, 15, "physical", "", false, false], [19, 20, 3, 15, "role", "", false, false], [19, 20, 3, 15, "temporal", "", false, false], [22, 23, 3, 15, "physical", "", false, false], [22, 23, 3, 15, "role", "", false, false], [22, 23, 3, 15, "temporal", "", false, false], [25, 26, 3, 15, "physical", "", false, false], [25, 26, 3, 15, "role", "", false, false], [25, 26, 3, 15, "temporal", "", false, false], [28, 30, 3, 15, "physical", "", false, false], [28, 30, 3, 15, "role", "", false, false], [28, 30, 3, 15, "temporal", "", false, false], [39, 40, 45, 48, "usage", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13], "sentence": ["Lors", "de", "la", "conf\u00e9rence", "de", "l'", "IEEE", "sur", "la", "vision", "informatique", "et", "la", "reconnaissance", "des", "formes", "en", "2006", ",", "Qiang", "Zhu", ",", "Shai", "Avidan", ",", "Mei-Chen", "Yeh", "et", "Kwang-Ting", "Cheng", "ont", "pr\u00e9sent\u00e9", "un", "algorithme", "permettant", "d'", "acc\u00e9l\u00e9rer", "consid\u00e9rablement", "la", "d\u00e9tection", "humaine", "\u00e0", "l'", "aide", "des", "m\u00e9thodes", "de", "descripteurs", "HOG", "."], "sentence-detokenized": "Lors de la conf\u00e9rence de l'IEEE sur la vision informatique et la reconnaissance des formes en 2006, Qiang Zhu, Shai Avidan, Mei-Chen Yeh et Kwang-Ting Cheng ont pr\u00e9sent\u00e9 un algorithme permettant d'acc\u00e9l\u00e9rer consid\u00e9rablement la d\u00e9tection humaine \u00e0 l'aide des m\u00e9thodes de descripteurs HOG.", "token2charspan": [[0, 4], [5, 7], [8, 10], [11, 21], [22, 24], [25, 27], [27, 31], [32, 35], [36, 38], [39, 45], [46, 58], [59, 61], [62, 64], [65, 79], [80, 83], [84, 90], [91, 93], [94, 98], [98, 99], [100, 105], [106, 109], [109, 110], [111, 115], [116, 122], [122, 123], [124, 132], [133, 136], [137, 139], [140, 150], [151, 156], [157, 160], [161, 169], [170, 172], [173, 183], [184, 194], [195, 197], [197, 206], [207, 223], [224, 226], [227, 236], [237, 244], [245, 246], [247, 249], [249, 253], [254, 257], [258, 266], [267, 269], [270, 282], [283, 286], [286, 287]]}
{"doc_key": "ai-dev-223", "ner": [[0, 1, "researcher"], [7, 7, "conference"], [11, 13, "organisation"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[0, 1, 7, 7, "role", "", false, false], [0, 1, 11, 13, "role", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["M.", "Hayes", "est", "membre", "fondateur", "de", "l'", "AAAI", "et", "de", "la", "Cognitive", "Science", "Society", "."], "sentence-detokenized": "M. Hayes est membre fondateur de l'AAAI et de la Cognitive Science Society.", "token2charspan": [[0, 2], [3, 8], [9, 12], [13, 19], [20, 29], [30, 32], [33, 35], [35, 39], [40, 42], [43, 45], [46, 48], [49, 58], [59, 66], [67, 74], [74, 75]]}
{"doc_key": "ai-dev-224", "ner": [[1, 2, "misc"], [7, 7, "field"], [10, 12, "field"], [15, 17, "field"], [20, 20, "field"], [23, 24, "field"], [27, 28, "field"], [31, 35, "field"], [38, 38, "field"], [41, 43, "field"], [46, 46, "field"], [49, 52, "field"], [62, 63, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12], "relations": [[1, 2, 7, 7, "part-of", "", false, false], [1, 2, 7, 7, "usage", "", false, false], [1, 2, 10, 12, "part-of", "", false, false], [1, 2, 10, 12, "usage", "", false, false], [1, 2, 15, 17, "part-of", "", false, false], [1, 2, 15, 17, "usage", "", false, false], [1, 2, 20, 20, "part-of", "", false, false], [1, 2, 20, 20, "usage", "", false, false], [1, 2, 23, 24, "part-of", "", false, false], [1, 2, 23, 24, "usage", "", false, false], [1, 2, 27, 28, "part-of", "", false, false], [1, 2, 27, 28, "usage", "", false, false], [1, 2, 31, 35, "part-of", "", false, false], [1, 2, 31, 35, "usage", "", false, false], [1, 2, 38, 38, "part-of", "", false, false], [1, 2, 38, 38, "usage", "", false, false], [1, 2, 41, 43, "part-of", "", false, false], [1, 2, 41, 43, "usage", "", false, false], [1, 2, 46, 46, "part-of", "", false, false], [1, 2, 46, 46, "usage", "", false, false], [1, 2, 49, 52, "part-of", "", false, false], [1, 2, 49, 52, "usage", "", false, false], [1, 2, 62, 63, "part-of", "", false, false], [1, 2, 62, 63, "usage", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23], "sentence": ["Les", "s\u00e9ries", "temporelles", "sont", "utilis\u00e9es", "dans", "les", "statistiques", ",", "le", "traitement", "des", "signaux", ",", "la", "reconnaissance", "des", "formes", ",", "l'", "\u00e9conom\u00e9trie", ",", "la", "finance", "math\u00e9matique", ",", "les", "pr\u00e9visions", "m\u00e9t\u00e9orologiques", ",", "la", "pr\u00e9vision", "des", "tremblements", "de", "terre", ",", "l'", "\u00e9lectroenc\u00e9phalographie", ",", "l'", "ing\u00e9nierie", "de", "contr\u00f4le", ",", "l'", "astronomie", ",", "l'", "ing\u00e9nierie", "des", "communications", "et", ",", "en", "g\u00e9n\u00e9ral", ",", "dans", "tous", "les", "domaines", "des", "sciences", "appliqu\u00e9es", "et", "de", "l'", "ing\u00e9nierie", "qui", "impliquent", "des", "mesures", "temporelles", "."], "sentence-detokenized": "Les s\u00e9ries temporelles sont utilis\u00e9es dans les statistiques, le traitement des signaux, la reconnaissance des formes, l'\u00e9conom\u00e9trie, la finance math\u00e9matique, les pr\u00e9visions m\u00e9t\u00e9orologiques, la pr\u00e9vision des tremblements de terre, l'\u00e9lectroenc\u00e9phalographie, l'ing\u00e9nierie de contr\u00f4le, l'astronomie, l'ing\u00e9nierie des communications et, en g\u00e9n\u00e9ral, dans tous les domaines des sciences appliqu\u00e9es et de l'ing\u00e9nierie qui impliquent des mesures temporelles.", "token2charspan": [[0, 3], [4, 10], [11, 22], [23, 27], [28, 37], [38, 42], [43, 46], [47, 59], [59, 60], [61, 63], [64, 74], [75, 78], [79, 86], [86, 87], [88, 90], [91, 105], [106, 109], [110, 116], [116, 117], [118, 120], [120, 131], [131, 132], [133, 135], [136, 143], [144, 156], [156, 157], [158, 161], [162, 172], [173, 188], [188, 189], [190, 192], [193, 202], [203, 206], [207, 219], [220, 222], [223, 228], [228, 229], [230, 232], [232, 255], [255, 256], [257, 259], [259, 269], [270, 272], [273, 281], [281, 282], [283, 285], [285, 295], [295, 296], [297, 299], [299, 309], [310, 313], [314, 328], [329, 331], [331, 332], [333, 335], [336, 343], [343, 344], [345, 349], [350, 354], [355, 358], [359, 367], [368, 371], [372, 380], [381, 391], [392, 394], [395, 397], [398, 400], [400, 410], [411, 414], [415, 425], [426, 429], [430, 437], [438, 449], [449, 450]]}
{"doc_key": "ai-dev-225", "ner": [[17, 19, "metrics"], [41, 41, "misc"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["En", "principe", ",", "la", "r\u00e9cup\u00e9ration", "exacte", "peut", "\u00eatre", "r\u00e9solue", "dans", "son", "domaine", "de", "faisabilit\u00e9", "en", "utilisant", "le", "maximum", "de", "vraisemblance", ",", "mais", "cela", "revient", "\u00e0", "r\u00e9soudre", "un", "probl\u00e8me", "de", "coupe", "contraint", "ou", "r\u00e9gularis\u00e9", "tel", "que", "la", "bissection", "minimale", "qui", "est", "typiquement", "NP-complet", "."], "sentence-detokenized": "En principe, la r\u00e9cup\u00e9ration exacte peut \u00eatre r\u00e9solue dans son domaine de faisabilit\u00e9 en utilisant le maximum de vraisemblance, mais cela revient \u00e0 r\u00e9soudre un probl\u00e8me de coupe contraint ou r\u00e9gularis\u00e9 tel que la bissection minimale qui est typiquement NP-complet.", "token2charspan": [[0, 2], [3, 11], [11, 12], [13, 15], [16, 28], [29, 35], [36, 40], [41, 45], [46, 53], [54, 58], [59, 62], [63, 70], [71, 73], [74, 85], [86, 88], [89, 98], [99, 101], [102, 109], [110, 112], [113, 126], [126, 127], [128, 132], [133, 137], [138, 145], [146, 147], [148, 156], [157, 159], [160, 168], [169, 171], [172, 177], [178, 187], [188, 190], [191, 201], [202, 205], [206, 209], [210, 212], [213, 223], [224, 232], [233, 236], [237, 240], [241, 252], [253, 263], [263, 264]]}
{"doc_key": "ai-dev-226", "ner": [[4, 6, "task"], [17, 17, "conference"]], "ner_mapping_to_source": [0, 1], "relations": [[17, 17, 4, 6, "topic", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["dans", "leur", "travail", "de", "d\u00e9tection", "des", "pi\u00e9tons", ",", "qui", "a", "\u00e9t\u00e9", "d\u00e9crit", "pour", "la", "premi\u00e8re", "fois", "au", "BMVC", "en", "2009", "."], "sentence-detokenized": "dans leur travail de d\u00e9tection des pi\u00e9tons, qui a \u00e9t\u00e9 d\u00e9crit pour la premi\u00e8re fois au BMVC en 2009.", "token2charspan": [[0, 4], [5, 9], [10, 17], [18, 20], [21, 30], [31, 34], [35, 42], [42, 43], [44, 47], [48, 49], [50, 53], [54, 60], [61, 65], [66, 68], [69, 77], [78, 82], [83, 85], [86, 90], [91, 93], [94, 98], [98, 99]]}
{"doc_key": "ai-dev-227", "ner": [[6, 12, "conference"], [14, 16, "researcher"], [19, 27, "misc"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[14, 16, 6, 12, "physical", "", false, false], [14, 16, 6, 12, "role", "", false, false], [14, 16, 19, 27, "win-defeat", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["En", "2007", ",", "lors", "de", "la", "conf\u00e9rence", "internationale", "sur", "la", "vision", "par", "ordinateur", ",", "M.", "Terzopoulos", "a", "re\u00e7u", "le", "premier", "prix", "IEEE", "PAMI", "Computer", "Vision", "Distinguished", "Researcher", "Award", "pour", "ses", "recherches", "pionni\u00e8res", "et", "soutenues", "sur", "les", "mod\u00e8les", "d\u00e9formables", "et", "leurs", "applications", "."], "sentence-detokenized": "En 2007, lors de la conf\u00e9rence internationale sur la vision par ordinateur, M. Terzopoulos a re\u00e7u le premier prix IEEE PAMI Computer Vision Distinguished Researcher Award pour ses recherches pionni\u00e8res et soutenues sur les mod\u00e8les d\u00e9formables et leurs applications.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 13], [14, 16], [17, 19], [20, 30], [31, 45], [46, 49], [50, 52], [53, 59], [60, 63], [64, 74], [74, 75], [76, 78], [79, 90], [91, 92], [93, 97], [98, 100], [101, 108], [109, 113], [114, 118], [119, 123], [124, 132], [133, 139], [140, 153], [154, 164], [165, 170], [171, 175], [176, 179], [180, 190], [191, 201], [202, 204], [205, 214], [215, 218], [219, 222], [223, 230], [231, 242], [243, 245], [246, 251], [252, 264], [264, 265]]}
{"doc_key": "ai-dev-228", "ner": [[0, 3, "task"], [5, 7, "task"]], "ner_mapping_to_source": [0, 1], "relations": [[0, 3, 5, 7, "named", "same", false, false]], "relations_mapping_to_source": [0], "sentence": ["L'", "analyse", "des", "clusters", "ou", "analyse", "par", "grappes", "consiste", "\u00e0", "attribuer", "des", "points", "de", "donn\u00e9es", "\u00e0", "des", "grappes", "de", "sorte", "que", "les", "\u00e9l\u00e9ments", "d'", "une", "m\u00eame", "grappe", "soient", "aussi", "semblables", "que", "possible", ",", "tandis", "que", "les", "\u00e9l\u00e9ments", "appartenant", "\u00e0", "des", "grappes", "diff\u00e9rentes", "sont", "aussi", "dissemblables", "que", "possible", "."], "sentence-detokenized": "L'analyse des clusters ou analyse par grappes consiste \u00e0 attribuer des points de donn\u00e9es \u00e0 des grappes de sorte que les \u00e9l\u00e9ments d'une m\u00eame grappe soient aussi semblables que possible, tandis que les \u00e9l\u00e9ments appartenant \u00e0 des grappes diff\u00e9rentes sont aussi dissemblables que possible.", "token2charspan": [[0, 2], [2, 9], [10, 13], [14, 22], [23, 25], [26, 33], [34, 37], [38, 45], [46, 54], [55, 56], [57, 66], [67, 70], [71, 77], [78, 80], [81, 88], [89, 90], [91, 94], [95, 102], [103, 105], [106, 111], [112, 115], [116, 119], [120, 128], [129, 131], [131, 134], [135, 139], [140, 146], [147, 153], [154, 159], [160, 170], [171, 174], [175, 183], [183, 184], [185, 191], [192, 195], [196, 199], [200, 208], [209, 220], [221, 222], [223, 226], [227, 234], [235, 246], [247, 251], [252, 257], [258, 271], [272, 275], [276, 284], [284, 285]]}
{"doc_key": "ai-dev-229", "ner": [[12, 14, "field"], [19, 21, "field"], [25, 27, "task"], [30, 32, "field"], [36, 39, "field"], [44, 46, "field"], [52, 54, "field"], [56, 58, "task"], [61, 63, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "relations": [[12, 14, 19, 21, "named", "", false, false], [12, 14, 30, 32, "named", "", false, false], [12, 14, 44, 46, "named", "", false, false], [25, 27, 19, 21, "part-of", "task_part_of_field", false, false], [36, 39, 30, 32, "part-of", "", false, false], [52, 54, 44, 46, "part-of", "", false, false], [56, 58, 52, 54, "part-of", "", false, false], [61, 63, 52, 54, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "sentence": ["(", "2005", ")", ",", "nous", "pouvons", "distinguer", "trois", "perspectives", "diff\u00e9rentes", "de", "l'", "exploration", "de", "texte", ",", "\u00e0", "savoir", "l'", "exploration", "de", "texte", "en", "tant", "qu'", "extraction", "d'", "information", ",", "l'", "exploration", "de", "texte", "en", "tant", "qu'", "exploration", "de", "donn\u00e9es", "de", "texte", ",", "et", "l'", "exploration", "de", "texte", "en", "tant", "que", "processus", "d'", "exploration", "de", "donn\u00e9es", "(", "d\u00e9couverte", "de", "connaissances", "dans", "des", "bases", "de", "donn\u00e9es", ")", ".Hotho", ",", "A.", ",", "N\u00fcrnberger", ",", "A.", "et", "Paa\u00df", ",", "G.", "(", "2005", ")", "."], "sentence-detokenized": "(2005), nous pouvons distinguer trois perspectives diff\u00e9rentes de l'exploration de texte, \u00e0 savoir l'exploration de texte en tant qu'extraction d'information, l'exploration de texte en tant qu'exploration de donn\u00e9es de texte, et l'exploration de texte en tant que processus d'exploration de donn\u00e9es (d\u00e9couverte de connaissances dans des bases de donn\u00e9es).Hotho, A., N\u00fcrnberger, A. et Paa\u00df, G. (2005).", "token2charspan": [[0, 1], [1, 5], [5, 6], [6, 7], [8, 12], [13, 20], [21, 31], [32, 37], [38, 50], [51, 62], [63, 65], [66, 68], [68, 79], [80, 82], [83, 88], [88, 89], [90, 91], [92, 98], [99, 101], [101, 112], [113, 115], [116, 121], [122, 124], [125, 129], [130, 133], [133, 143], [144, 146], [146, 157], [157, 158], [159, 161], [161, 172], [173, 175], [176, 181], [182, 184], [185, 189], [190, 193], [193, 204], [205, 207], [208, 215], [216, 218], [219, 224], [224, 225], [226, 228], [229, 231], [231, 242], [243, 245], [246, 251], [252, 254], [255, 259], [260, 263], [264, 273], [274, 276], [276, 287], [288, 290], [291, 298], [299, 300], [300, 310], [311, 313], [314, 327], [328, 332], [333, 336], [337, 342], [343, 345], [346, 353], [353, 354], [354, 360], [360, 361], [362, 364], [364, 365], [366, 376], [376, 377], [378, 380], [381, 383], [384, 388], [388, 389], [390, 392], [393, 394], [394, 398], [398, 399], [399, 400]]}
{"doc_key": "ai-dev-230", "ner": [[1, 2, "product"], [16, 22, "location"], [24, 24, "location"], [27, 27, "location"], [39, 41, "university"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[1, 2, 16, 22, "related-to", "developed_for", false, false], [16, 22, 24, 24, "physical", "", false, false], [24, 24, 27, 27, "physical", "", false, false], [39, 41, 1, 2, "role", "buys", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Le", "Rancho", "Arm", "a", "\u00e9t\u00e9", "d\u00e9velopp\u00e9", "comme", "un", "bras", "robotique", "pour", "aider", "les", "patients", "handicap\u00e9s", "du", "centre", "national", "de", "r\u00e9habilitation", "Rancho", "Los", "Amigos", "\u00e0", "Downey", ",", "en", "Californie", ";", "ce", "bras", "contr\u00f4l\u00e9", "par", "ordinateur", "a", "\u00e9t\u00e9", "achet\u00e9", "par", "l'", "universit\u00e9", "de", "Stanford", "en", "1963", "."], "sentence-detokenized": "Le Rancho Arm a \u00e9t\u00e9 d\u00e9velopp\u00e9 comme un bras robotique pour aider les patients handicap\u00e9s du centre national de r\u00e9habilitation Rancho Los Amigos \u00e0 Downey, en Californie ; ce bras contr\u00f4l\u00e9 par ordinateur a \u00e9t\u00e9 achet\u00e9 par l'universit\u00e9 de Stanford en 1963.", "token2charspan": [[0, 2], [3, 9], [10, 13], [14, 15], [16, 19], [20, 29], [30, 35], [36, 38], [39, 43], [44, 53], [54, 58], [59, 64], [65, 68], [69, 77], [78, 88], [89, 91], [92, 98], [99, 107], [108, 110], [111, 125], [126, 132], [133, 136], [137, 143], [144, 145], [146, 152], [152, 153], [154, 156], [157, 167], [168, 169], [170, 172], [173, 177], [178, 186], [187, 190], [191, 201], [202, 203], [204, 207], [208, 214], [215, 218], [219, 221], [221, 231], [232, 234], [235, 243], [244, 246], [247, 251], [251, 252]]}
{"doc_key": "ai-dev-231", "ner": [[2, 2, "university"], [4, 5, "researcher"], [12, 16, "organisation"], [24, 26, "organisation"], [29, 30, "researcher"], [32, 34, "researcher"], [51, 51, "university"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[4, 5, 2, 2, "physical", "", false, false], [4, 5, 2, 2, "role", "", false, false], [4, 5, 12, 16, "role", "founder", false, false], [4, 5, 24, 26, "role", "founder", false, false], [24, 26, 51, 51, "physical", "", false, false], [29, 30, 24, 26, "role", "founder", false, false], [32, 34, 24, 26, "role", "founder", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "sentence": ["\u00c0", "l'", "UCSD", ",", "Norman", "a", "\u00e9t\u00e9", "l'", "un", "des", "fondateurs", "de", "l'", "Institute", "for", "Cognitive", "Science", "et", "l'", "un", "des", "organisateurs", "de", "la", "Cognitive", "Science", "Society", "(", "avec", "Roger", "Schank", ",", "Allan", "M.", "Collins", "et", "d'", "autres", ")", ",", "qui", "a", "tenu", "sa", "premi\u00e8re", "r\u00e9union", "sur", "le", "campus", "de", "l'", "UCSD", "en", "1979", "."], "sentence-detokenized": "\u00c0 l'UCSD, Norman a \u00e9t\u00e9 l'un des fondateurs de l'Institute for Cognitive Science et l'un des organisateurs de la Cognitive Science Society (avec Roger Schank, Allan M. Collins et d'autres), qui a tenu sa premi\u00e8re r\u00e9union sur le campus de l'UCSD en 1979.", "token2charspan": [[0, 1], [2, 4], [4, 8], [8, 9], [10, 16], [17, 18], [19, 22], [23, 25], [25, 27], [28, 31], [32, 42], [43, 45], [46, 48], [48, 57], [58, 61], [62, 71], [72, 79], [80, 82], [83, 85], [85, 87], [88, 91], [92, 105], [106, 108], [109, 111], [112, 121], [122, 129], [130, 137], [138, 139], [139, 143], [144, 149], [150, 156], [156, 157], [158, 163], [164, 166], [167, 174], [175, 177], [178, 180], [180, 186], [186, 187], [187, 188], [189, 192], [193, 194], [195, 199], [200, 202], [203, 211], [212, 219], [220, 223], [224, 226], [227, 233], [234, 236], [237, 239], [239, 243], [244, 246], [247, 251], [251, 252]]}
{"doc_key": "ai-dev-232", "ner": [[10, 11, "product"], [14, 15, "product"], [18, 19, "product"], [22, 25, "product"], [27, 28, "product"], [30, 31, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[27, 28, 22, 25, "type-of", "", false, false], [30, 31, 22, 25, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Les", "configurations", "de", "robots", "les", "plus", "couramment", "utilis\u00e9es", "sont", "les", "robots", "articul\u00e9s", ",", "les", "robots", "SCARA", ",", "les", "robots", "delta", "et", "les", "robots", "\u00e0", "coordonn\u00e9es", "cart\u00e9siennes", "(", "robots", "portiques", "ou", "robots", "x-y-z", ")", "."], "sentence-detokenized": "Les configurations de robots les plus couramment utilis\u00e9es sont les robots articul\u00e9s, les robots SCARA, les robots delta et les robots \u00e0 coordonn\u00e9es cart\u00e9siennes (robots portiques ou robots x-y-z).", "token2charspan": [[0, 3], [4, 18], [19, 21], [22, 28], [29, 32], [33, 37], [38, 48], [49, 58], [59, 63], [64, 67], [68, 74], [75, 84], [84, 85], [86, 89], [90, 96], [97, 102], [102, 103], [104, 107], [108, 114], [115, 120], [121, 123], [124, 127], [128, 134], [135, 136], [137, 148], [149, 161], [162, 163], [163, 169], [170, 179], [180, 182], [183, 189], [190, 195], [195, 196], [196, 197]]}
{"doc_key": "ai-dev-233", "ner": [[9, 9, "programlang"], [8, 10, "misc"], [17, 17, "misc"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[8, 10, 9, 9, "part-of", "", false, false], [17, 17, 8, 10, "related-to", "compatible_with", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Il", "peut", "\u00e9galement", "\u00eatre", "utilis\u00e9", "directement", "avec", "le", "module", "Perl", "TM", "(", "qui", "prend", "\u00e9galement", "en", "charge", "LTM", ")", "."], "sentence-detokenized": "Il peut \u00e9galement \u00eatre utilis\u00e9 directement avec le module Perl TM (qui prend \u00e9galement en charge LTM).", "token2charspan": [[0, 2], [3, 7], [8, 17], [18, 22], [23, 30], [31, 42], [43, 47], [48, 50], [51, 57], [58, 62], [63, 65], [66, 67], [67, 70], [71, 76], [77, 86], [87, 89], [90, 96], [97, 100], [100, 101], [101, 102]]}
{"doc_key": "ai-dev-234", "ner": [[10, 11, "organisation"], [17, 17, "organisation"]], "ner_mapping_to_source": [1, 2], "relations": [], "relations_mapping_to_source": [], "sentence": ["La", "comp\u00e9tition", "a", "\u00e9t\u00e9", "remport\u00e9e", "par", "une", "\u00e9quipe", "am\u00e9ricaine", "de", "Newton", "Labs", "et", "a", "\u00e9t\u00e9", "diffus\u00e9e", "sur", "CNN", "."], "sentence-detokenized": "La comp\u00e9tition a \u00e9t\u00e9 remport\u00e9e par une \u00e9quipe am\u00e9ricaine de Newton Labs et a \u00e9t\u00e9 diffus\u00e9e sur CNN.", "token2charspan": [[0, 2], [3, 14], [15, 16], [17, 20], [21, 30], [31, 34], [35, 38], [39, 45], [46, 56], [57, 59], [60, 66], [67, 71], [72, 74], [75, 76], [77, 80], [81, 89], [90, 93], [94, 97], [97, 98]]}
{"doc_key": "ai-dev-235", "ner": [[0, 3, "misc"], [9, 10, "person"], [12, 13, "person"], [15, 18, "person"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[9, 10, 0, 3, "role", "directs", false, false], [12, 13, 0, 3, "role", "acts_in", false, false], [15, 18, 0, 3, "role", "acts_in", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["The", "Butler's", "in", "Love", ",", "un", "court-m\u00e9trage", "r\u00e9alis\u00e9", "par", "David", "Arquette", "avec", "Elizabeth", "Berkley", "et", "Thomas", "Jane", ",", "est", "sorti", "le", "23", "juin", "2008", "."], "sentence-detokenized": "The Butler's in Love, un court-m\u00e9trage r\u00e9alis\u00e9 par David Arquette avec Elizabeth Berkley et Thomas Jane, est sorti le 23 juin 2008.", "token2charspan": [[0, 3], [4, 12], [13, 15], [16, 20], [20, 21], [22, 24], [25, 38], [39, 46], [47, 50], [51, 56], [57, 65], [66, 70], [71, 80], [81, 88], [89, 91], [92, 98], [99, 103], [103, 104], [105, 108], [109, 114], [115, 117], [118, 120], [121, 125], [126, 130], [130, 131]]}
{"doc_key": "ai-dev-236", "ner": [[3, 3, "product"], [9, 9, "field"], [19, 19, "misc"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[3, 3, 19, 19, "general-affiliation", "", false, false], [9, 9, 3, 3, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Par", "exemple", ",", "WordNet", "est", "une", "ressource", "comprenant", "une", "taxonomie", ",", "dont", "les", "\u00e9l\u00e9ments", "sont", "les", "significations", "des", "mots", "anglais", "."], "sentence-detokenized": "Par exemple, WordNet est une ressource comprenant une taxonomie, dont les \u00e9l\u00e9ments sont les significations des mots anglais.", "token2charspan": [[0, 3], [4, 11], [11, 12], [13, 20], [21, 24], [25, 28], [29, 38], [39, 49], [50, 53], [54, 63], [63, 64], [65, 69], [70, 73], [74, 82], [83, 87], [88, 91], [92, 106], [107, 110], [111, 115], [116, 123], [123, 124]]}
{"doc_key": "ai-dev-237", "ner": [[0, 4, "product"], [9, 9, "product"], [11, 11, "product"], [20, 20, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[9, 9, 0, 4, "type-of", "", false, false], [9, 9, 20, 20, "related-to", "ability_to", false, false], [11, 11, 0, 4, "type-of", "", false, false], [11, 11, 20, 20, "related-to", "ability_to", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Les", "syst\u00e8mes", "de", "robots", "humano\u00efdes", "existants", ",", "tels", "que", "ASIMO", "et", "QRIO", ",", "utilisent", "de", "nombreux", "moteurs", "pour", "assurer", "la", "locomotion", "."], "sentence-detokenized": "Les syst\u00e8mes de robots humano\u00efdes existants, tels que ASIMO et QRIO, utilisent de nombreux moteurs pour assurer la locomotion.", "token2charspan": [[0, 3], [4, 12], [13, 15], [16, 22], [23, 33], [34, 43], [43, 44], [45, 49], [50, 53], [54, 59], [60, 62], [63, 67], [67, 68], [69, 78], [79, 81], [82, 90], [91, 98], [99, 103], [104, 111], [112, 114], [115, 125], [125, 126]]}
{"doc_key": "ai-dev-238", "ner": [[0, 0, "metrics"], [7, 9, "metrics"], [13, 13, "metrics"], [16, 22, "misc"], [25, 25, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[7, 9, 0, 0, "part-of", "", false, false], [13, 13, 0, 0, "part-of", "", false, false], [16, 22, 0, 0, "part-of", "", false, false], [25, 25, 0, 0, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["LEPOR", "est", "con\u00e7u", "avec", "les", "facteurs", "de", "p\u00e9nalit\u00e9", "de", "longueur", "am\u00e9lior\u00e9e", ",", "de", "pr\u00e9cision", ",", "de", "p\u00e9nalit\u00e9", "d'", "ordre", "de", "mots", "n-", "grammes", "et", "de", "rappel", "."], "sentence-detokenized": "LEPOR est con\u00e7u avec les facteurs de p\u00e9nalit\u00e9 de longueur am\u00e9lior\u00e9e, de pr\u00e9cision, de p\u00e9nalit\u00e9 d'ordre de mots n-grammes et de rappel.", "token2charspan": [[0, 5], [6, 9], [10, 15], [16, 20], [21, 24], [25, 33], [34, 36], [37, 45], [46, 48], [49, 57], [58, 67], [67, 68], [69, 71], [72, 81], [81, 82], [83, 85], [86, 94], [95, 97], [97, 102], [103, 105], [106, 110], [111, 113], [113, 120], [121, 123], [124, 126], [127, 133], [133, 134]]}
{"doc_key": "ai-dev-239", "ner": [[5, 9, "metrics"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "est", "bas\u00e9", "sur", "la", "m\u00e9trique", "de", "l'", "\u00e9valuation", "bilingue", ",", "mais", "avec", "quelques", "modifications", "."], "sentence-detokenized": "Il est bas\u00e9 sur la m\u00e9trique de l'\u00e9valuation bilingue, mais avec quelques modifications.", "token2charspan": [[0, 2], [3, 6], [7, 11], [12, 15], [16, 18], [19, 27], [28, 30], [31, 33], [33, 43], [44, 52], [52, 53], [54, 58], [59, 63], [64, 72], [73, 86], [86, 87]]}
{"doc_key": "ai-dev-240", "ner": [[11, 11, "product"], [13, 13, "product"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "s'", "agit", "d'", "un", "exemple", "de", "mise", "en", "\u0153uvre", "dans", "MATLAB", "/", "Octave", ":"], "sentence-detokenized": "Il s'agit d'un exemple de mise en \u0153uvre dans MATLAB / Octave :", "token2charspan": [[0, 2], [3, 5], [5, 9], [10, 12], [12, 14], [15, 22], [23, 25], [26, 30], [31, 33], [34, 39], [40, 44], [45, 51], [52, 53], [54, 60], [61, 62]]}
{"doc_key": "ai-dev-241", "ner": [[15, 15, "programlang"], [17, 17, "programlang"], [19, 19, "programlang"]], "ner_mapping_to_source": [0, 1, 2], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "est", "con\u00e7u", "pour", "\u00eatre", "utilis\u00e9", "par", "un", "certain", "nombre", "de", "langages", "informatiques", ",", "dont", "Python", ",", "Ruby", "et", "Scheme", "."], "sentence-detokenized": "Il est con\u00e7u pour \u00eatre utilis\u00e9 par un certain nombre de langages informatiques, dont Python, Ruby et Scheme.", "token2charspan": [[0, 2], [3, 6], [7, 12], [13, 17], [18, 22], [23, 30], [31, 34], [35, 37], [38, 45], [46, 52], [53, 55], [56, 64], [65, 78], [78, 79], [80, 84], [85, 91], [91, 92], [93, 97], [98, 100], [101, 107], [107, 108]]}
{"doc_key": "ai-dev-242", "ner": [[0, 1, "researcher"], [7, 7, "organisation"], [14, 14, "conference"], [19, 20, "academicjournal"], [25, 27, "organisation"], [32, 36, "organisation"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[0, 1, 7, 7, "role", "", false, false], [0, 1, 14, 14, "role", "", false, false], [0, 1, 19, 20, "role", "", false, false], [0, 1, 25, 27, "role", "", false, false], [0, 1, 32, 36, "role", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["M.", "Hayes", "a", "\u00e9t\u00e9", "secr\u00e9taire", "de", "l'", "AISB", ",", "pr\u00e9sident", "et", "administrateur", "de", "l'", "IJCAI", ",", "r\u00e9dacteur", "associ\u00e9", "d'", "Artificial", "Intelligence", ",", "gouverneur", "de", "la", "Cognitive", "Science", "Society", "et", "pr\u00e9sident", "de", "l'", "American", "Association", "for", "Artificial", "Intelligence", "."], "sentence-detokenized": "M. Hayes a \u00e9t\u00e9 secr\u00e9taire de l'AISB, pr\u00e9sident et administrateur de l'IJCAI, r\u00e9dacteur associ\u00e9 d'Artificial Intelligence, gouverneur de la Cognitive Science Society et pr\u00e9sident de l'American Association for Artificial Intelligence.", "token2charspan": [[0, 2], [3, 8], [9, 10], [11, 14], [15, 25], [26, 28], [29, 31], [31, 35], [35, 36], [37, 46], [47, 49], [50, 64], [65, 67], [68, 70], [70, 75], [75, 76], [77, 86], [87, 94], [95, 97], [97, 107], [108, 120], [120, 121], [122, 132], [133, 135], [136, 138], [139, 148], [149, 156], [157, 164], [165, 167], [168, 177], [178, 180], [181, 183], [183, 191], [192, 203], [204, 207], [208, 218], [219, 231], [231, 232]]}
{"doc_key": "ai-dev-243", "ner": [[5, 15, "misc"], [17, 19, "misc"], [25, 26, "person"], [31, 36, "organisation"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[25, 26, 5, 15, "role", "directed_by", false, false], [25, 26, 17, 19, "role", "directed_by", false, false], [25, 26, 31, 36, "role", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Deux", "d'", "entre", "eux", ",", "Now", "is", "the", "Time", "(", "to", "Put", "On", "Your", "Glasses", ")", "et", "Around", "is", "Around", ",", "ont", "\u00e9t\u00e9", "r\u00e9alis\u00e9s", "par", "Norman", "McLaren", "en", "1951", "pour", "l'", "Office", "national", "du", "film", "du", "Canada", "."], "sentence-detokenized": "Deux d'entre eux, Now is the Time (to Put On Your Glasses) et Around is Around, ont \u00e9t\u00e9 r\u00e9alis\u00e9s par Norman McLaren en 1951 pour l'Office national du film du Canada.", "token2charspan": [[0, 4], [5, 7], [7, 12], [13, 16], [16, 17], [18, 21], [22, 24], [25, 28], [29, 33], [34, 35], [35, 37], [38, 41], [42, 44], [45, 49], [50, 57], [57, 58], [59, 61], [62, 68], [69, 71], [72, 78], [78, 79], [80, 83], [84, 87], [88, 96], [97, 100], [101, 107], [108, 115], [116, 118], [119, 123], [124, 128], [129, 131], [131, 137], [138, 146], [147, 149], [150, 154], [155, 157], [158, 164], [164, 165]]}
{"doc_key": "ai-dev-244", "ner": [[1, 3, "product"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Un", "syst\u00e8me", "de", "recommandation", "vise", "\u00e0", "pr\u00e9dire", "la", "pr\u00e9f\u00e9rence", "pour", "un", "article", "d'", "un", "utilisateur", "cible", "."], "sentence-detokenized": "Un syst\u00e8me de recommandation vise \u00e0 pr\u00e9dire la pr\u00e9f\u00e9rence pour un article d'un utilisateur cible.", "token2charspan": [[0, 2], [3, 10], [11, 13], [14, 28], [29, 33], [34, 35], [36, 43], [44, 46], [47, 57], [58, 62], [63, 65], [66, 73], [74, 76], [76, 78], [79, 90], [91, 96], [96, 97]]}
{"doc_key": "ai-dev-245", "ner": [[0, 1, "algorithm"], [10, 10, "field"], [12, 12, "field"], [14, 16, "field"], [18, 21, "field"], [23, 26, "field"], [28, 29, "field"], [31, 31, "field"], [33, 34, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "relations": [[0, 1, 10, 10, "part-of", "", true, false], [0, 1, 12, 12, "part-of", "", true, false], [0, 1, 14, 16, "part-of", "", true, false], [0, 1, 18, 21, "part-of", "", true, false], [0, 1, 23, 26, "part-of", "", true, false], [0, 1, 28, 29, "part-of", "", true, false], [0, 1, 31, 31, "part-of", "", true, false], [0, 1, 33, 34, "part-of", "", true, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "sentence": ["La", "convolution", "trouve", "des", "applications", "dans", "les", "domaines", "suivants", ":", "probabilit\u00e9s", ",", "statistiques", ",", "vision", "par", "ordinateur", ",", "traitement", "du", "langage", "naturel", ",", "traitement", "de", "l'", "image", "et", "du", "signal", ",", "ing\u00e9nierie", "et", "\u00e9quations", "diff\u00e9rentielles", "."], "sentence-detokenized": "La convolution trouve des applications dans les domaines suivants : probabilit\u00e9s, statistiques, vision par ordinateur, traitement du langage naturel, traitement de l'image et du signal, ing\u00e9nierie et \u00e9quations diff\u00e9rentielles.", "token2charspan": [[0, 2], [3, 14], [15, 21], [22, 25], [26, 38], [39, 43], [44, 47], [48, 56], [57, 65], [66, 67], [68, 80], [80, 81], [82, 94], [94, 95], [96, 102], [103, 106], [107, 117], [117, 118], [119, 129], [130, 132], [133, 140], [141, 148], [148, 149], [150, 160], [161, 163], [164, 166], [166, 171], [172, 174], [175, 177], [178, 184], [184, 185], [186, 196], [197, 199], [200, 209], [210, 225], [225, 226]]}
{"doc_key": "ai-dev-246", "ner": [[3, 3, "field"], [6, 9, "task"], [12, 13, "task"], [19, 19, "task"], [18, 18, "task"], [16, 17, "task"], [22, 23, "task"], [26, 29, "task"], [32, 33, "task"], [36, 37, "task"], [40, 41, "task"], [44, 44, "field"], [47, 47, "field"], [50, 53, "field"], [56, 56, "field"], [59, 59, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15], "relations": [[3, 3, 6, 9, "part-of", "", true, false], [3, 3, 12, 13, "part-of", "", true, false], [3, 3, 19, 19, "part-of", "", true, false], [3, 3, 18, 18, "part-of", "", true, false], [3, 3, 16, 17, "part-of", "", true, false], [3, 3, 22, 23, "part-of", "", true, false], [3, 3, 26, 29, "part-of", "", true, false], [3, 3, 32, 33, "part-of", "", true, false], [3, 3, 36, 37, "part-of", "", true, false], [3, 3, 40, 41, "part-of", "", true, false], [3, 3, 44, 44, "part-of", "", true, false], [3, 3, 47, 47, "part-of", "", true, false], [3, 3, 50, 53, "part-of", "", true, false], [3, 3, 56, 56, "part-of", "", true, false], [3, 3, 59, 59, "part-of", "", true, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14], "sentence": ["Les", "applications", "du", "DSP", "comprennent", "le", "traitement", "des", "signaux", "audio", ",", "la", "compression", "audio", ",", "le", "traitement", "des", "images", "num\u00e9riques", ",", "la", "compression", "vid\u00e9o", ",", "le", "traitement", "de", "la", "parole", ",", "la", "reconnaissance", "vocale", ",", "les", "communications", "num\u00e9riques", ",", "les", "synth\u00e9tiseurs", "num\u00e9riques", ",", "le", "radar", ",", "le", "sonar", ",", "le", "traitement", "des", "signaux", "financiers", ",", "la", "sismologie", "et", "la", "biom\u00e9decine", "."], "sentence-detokenized": "Les applications du DSP comprennent le traitement des signaux audio, la compression audio, le traitement des images num\u00e9riques, la compression vid\u00e9o, le traitement de la parole, la reconnaissance vocale, les communications num\u00e9riques, les synth\u00e9tiseurs num\u00e9riques, le radar, le sonar, le traitement des signaux financiers, la sismologie et la biom\u00e9decine.", "token2charspan": [[0, 3], [4, 16], [17, 19], [20, 23], [24, 35], [36, 38], [39, 49], [50, 53], [54, 61], [62, 67], [67, 68], [69, 71], [72, 83], [84, 89], [89, 90], [91, 93], [94, 104], [105, 108], [109, 115], [116, 126], [126, 127], [128, 130], [131, 142], [143, 148], [148, 149], [150, 152], [153, 163], [164, 166], [167, 169], [170, 176], [176, 177], [178, 180], [181, 195], [196, 202], [202, 203], [204, 207], [208, 222], [223, 233], [233, 234], [235, 238], [239, 252], [253, 263], [263, 264], [265, 267], [268, 273], [273, 274], [275, 277], [278, 283], [283, 284], [285, 287], [288, 298], [299, 302], [303, 310], [311, 321], [321, 322], [323, 325], [326, 336], [337, 339], [340, 342], [343, 354], [354, 355]]}
{"doc_key": "ai-dev-247", "ner": [[11, 12, "misc"], [19, 19, "product"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["(", "20", "f\u00e9vrier", "1912", "-", "11", "ao\u00fbt", "2011", ")", "est", "un", "inventeur", "am\u00e9ricain", ",", "principalement", "connu", "pour", "avoir", "cr\u00e9\u00e9", "Unimate", ",", "le", "premier", "robot", "industriel", "."], "sentence-detokenized": "(20 f\u00e9vrier 1912 - 11 ao\u00fbt 2011) est un inventeur am\u00e9ricain, principalement connu pour avoir cr\u00e9\u00e9 Unimate, le premier robot industriel.", "token2charspan": [[0, 1], [1, 3], [4, 11], [12, 16], [17, 18], [19, 21], [22, 26], [27, 31], [31, 32], [33, 36], [37, 39], [40, 49], [50, 59], [59, 60], [61, 75], [76, 81], [82, 86], [87, 92], [93, 97], [98, 105], [105, 106], [107, 109], [110, 117], [118, 123], [124, 134], [134, 135]]}
{"doc_key": "ai-dev-248", "ner": [[1, 3, "researcher"], [5, 7, "researcher"], [9, 9, "researcher"], [28, 30, "algorithm"], [35, 37, "algorithm"], [44, 46, "task"], [49, 49, "algorithm"], [54, 55, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "relations": [[1, 3, 28, 30, "related-to", "writes_about", true, false], [5, 7, 28, 30, "related-to", "writes_about", true, false], [9, 9, 28, 30, "related-to", "writes_about", true, false], [28, 30, 35, 37, "related-to", "", true, false], [44, 46, 49, 49, "related-to", "", true, false], [54, 55, 49, 49, "origin", "", false, true]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["Avec", "David", "E.", "Rumelhart", "et", "Ronald", "J.", "Williams", ",", "Hinton", "a", "\u00e9t\u00e9", "le", "co-auteur", "d'", "un", "article", "tr\u00e8s", "cit\u00e9", ",", "publi\u00e9", "en", "1986", ",", "qui", "a", "popularis\u00e9", "l'", "algorithme", "de", "r\u00e9tropropagation", "pour", "la", "formation", "de", "r\u00e9seaux", "neuronaux", "multicouches", ".", "L'", "\u00e9tape", "spectaculaire", "de", "la", "reconnaissance", "d'", "images", "de", "l'", "AlexNet", "con\u00e7u", "par", "son", "\u00e9tudiant", "Alex", "Krizhevsky", "{{", "cite", "web"], "sentence-detokenized": "Avec David E. Rumelhart et Ronald J. Williams, Hinton a \u00e9t\u00e9 le co-auteur d'un article tr\u00e8s cit\u00e9, publi\u00e9 en 1986, qui a popularis\u00e9 l'algorithme de r\u00e9tropropagation pour la formation de r\u00e9seaux neuronaux multicouches. L'\u00e9tape spectaculaire de la reconnaissance d'images de l'AlexNet con\u00e7u par son \u00e9tudiant Alex Krizhevsky {{cite web", "token2charspan": [[0, 4], [5, 10], [11, 13], [14, 23], [24, 26], [27, 33], [34, 36], [37, 45], [45, 46], [47, 53], [54, 55], [56, 59], [60, 62], [63, 72], [73, 75], [75, 77], [78, 85], [86, 90], [91, 95], [95, 96], [97, 103], [104, 106], [107, 111], [111, 112], [113, 116], [117, 118], [119, 129], [130, 132], [132, 142], [143, 145], [146, 162], [163, 167], [168, 170], [171, 180], [181, 183], [184, 191], [192, 201], [202, 214], [214, 215], [216, 218], [218, 223], [224, 237], [238, 240], [241, 243], [244, 258], [259, 261], [261, 267], [268, 270], [271, 273], [273, 280], [281, 286], [287, 290], [291, 294], [295, 303], [304, 308], [309, 319], [320, 322], [322, 326], [327, 330]]}
{"doc_key": "ai-dev-249", "ner": [[12, 14, "metrics"], [17, 20, "metrics"], [23, 25, "metrics"]], "ner_mapping_to_source": [0, 1, 2], "relations": [], "relations_mapping_to_source": [], "sentence": ["Lorsque", "la", "valeur", "\u00e0", "pr\u00e9dire", "est", "distribu\u00e9e", "de", "mani\u00e8re", "continue", ",", "l'", "erreur", "quadratique", "moyenne", ",", "l'", "erreur", "quadratique", "moyenne", "racine", "ou", "l'", "\u00e9cart", "absolu", "m\u00e9dian", "peuvent", "\u00eatre", "utilis\u00e9s", "pour", "r\u00e9sumer", "les", "erreurs", "."], "sentence-detokenized": "Lorsque la valeur \u00e0 pr\u00e9dire est distribu\u00e9e de mani\u00e8re continue, l'erreur quadratique moyenne, l'erreur quadratique moyenne racine ou l'\u00e9cart absolu m\u00e9dian peuvent \u00eatre utilis\u00e9s pour r\u00e9sumer les erreurs.", "token2charspan": [[0, 7], [8, 10], [11, 17], [18, 19], [20, 27], [28, 31], [32, 42], [43, 45], [46, 53], [54, 62], [62, 63], [64, 66], [66, 72], [73, 84], [85, 92], [92, 93], [94, 96], [96, 102], [103, 114], [115, 122], [123, 129], [130, 132], [133, 135], [135, 140], [141, 147], [148, 154], [155, 162], [163, 167], [168, 176], [177, 181], [182, 189], [190, 193], [194, 201], [201, 202]]}
{"doc_key": "ai-dev-250", "ner": [[0, 2, "algorithm"], [18, 19, "field"], [22, 24, "field"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[0, 2, 18, 19, "part-of", "", true, false], [0, 2, 22, 24, "part-of", "", true, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Le", "clustering", "conceptuel", "s'", "est", "principalement", "d\u00e9velopp\u00e9", "au", "cours", "des", "ann\u00e9es", "1980", ",", "en", "tant", "que", "paradigme", "d'", "apprentissage", "automatique", "pour", "l'", "apprentissage", "non", "supervis\u00e9", "."], "sentence-detokenized": "Le clustering conceptuel s'est principalement d\u00e9velopp\u00e9 au cours des ann\u00e9es 1980, en tant que paradigme d'apprentissage automatique pour l'apprentissage non supervis\u00e9.", "token2charspan": [[0, 2], [3, 13], [14, 24], [25, 27], [27, 30], [31, 45], [46, 55], [56, 58], [59, 64], [65, 68], [69, 75], [76, 80], [80, 81], [82, 84], [85, 89], [90, 93], [94, 103], [104, 106], [106, 119], [120, 131], [132, 136], [137, 139], [139, 152], [153, 156], [157, 166], [166, 167]]}
{"doc_key": "ai-dev-251", "ner": [[11, 12, "product"], [35, 40, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Si", "les", "entit\u00e9s", "nomm\u00e9es", "ne", "peuvent", "pas", "\u00eatre", "reconnues", "par", "le", "traducteur", "automatique", ",", "elles", "peuvent", "\u00eatre", "traduites", "par", "erreur", "comme", "des", "noms", "communs", ",", "ce", "qui", "n'", "affecterait", "probablement", "pas", "la", "note", "de", "la", "doublure", "de", "l'", "\u00e9valuation", "bilingue", "de", "la", "traduction", "mais", "modifierait", "la", "lisibilit\u00e9", "humaine", "du", "texte", "."], "sentence-detokenized": "Si les entit\u00e9s nomm\u00e9es ne peuvent pas \u00eatre reconnues par le traducteur automatique, elles peuvent \u00eatre traduites par erreur comme des noms communs, ce qui n'affecterait probablement pas la note de la doublure de l'\u00e9valuation bilingue de la traduction mais modifierait la lisibilit\u00e9 humaine du texte.", "token2charspan": [[0, 2], [3, 6], [7, 14], [15, 22], [23, 25], [26, 33], [34, 37], [38, 42], [43, 52], [53, 56], [57, 59], [60, 70], [71, 82], [82, 83], [84, 89], [90, 97], [98, 102], [103, 112], [113, 116], [117, 123], [124, 129], [130, 133], [134, 138], [139, 146], [146, 147], [148, 150], [151, 154], [155, 157], [157, 168], [169, 181], [182, 185], [186, 188], [189, 193], [194, 196], [197, 199], [200, 208], [209, 211], [212, 214], [214, 224], [225, 233], [234, 236], [237, 239], [240, 250], [251, 255], [256, 267], [268, 270], [271, 281], [282, 289], [290, 292], [293, 298], [298, 299]]}
{"doc_key": "ai-dev-252", "ner": [[0, 1, "researcher"], [10, 11, "misc"], [12, 18, "conference"], [20, 20, "location"], [22, 22, "country"], [35, 36, "researcher"], [46, 46, "researcher"], [49, 51, "university"], [55, 56, "researcher"], [58, 59, "researcher"], [61, 62, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10], "relations": [[0, 1, 10, 11, "related-to", "writes_about", false, false], [0, 1, 12, 18, "temporal", "", true, false], [12, 18, 20, 20, "physical", "", false, false], [20, 20, 22, 22, "physical", "", false, false], [46, 46, 49, 51, "physical", "", false, false], [46, 46, 49, 51, "role", "", false, false], [55, 56, 49, 51, "physical", "", false, false], [55, 56, 49, 51, "role", "", false, false], [58, 59, 49, 51, "physical", "", false, false], [58, 59, 49, 51, "role", "", false, false], [61, 62, 49, 51, "physical", "", false, false], [61, 62, 49, 51, "role", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], "sentence": ["Roger", "Schank", ",", "1969", ",", "A", "conceptual", "dependency", "parser", "for", "natural", "language", "Proceedings", "of", "the", "1969", "on", "Computational", "linguistics", ",", "S\u00e5ng-S\u00e4by", ",", "Sweden", ",", "pages", "1-3", "Ce", "mod\u00e8le", ",", "partiellement", "influenc\u00e9", "par", "les", "travaux", "de", "Sydney", "Lamb", ",", "a", "\u00e9t\u00e9", "largement", "utilis\u00e9", "par", "les", "\u00e9tudiants", "de", "Schank", "\u00e0", "l'", "universit\u00e9", "de", "Yale", ",", "tels", "que", "Robert", "Wilensky", ",", "Wendy", "Lehnert", "et", "Janet", "Kolodner", "."], "sentence-detokenized": "Roger Schank, 1969, A conceptual dependency parser for natural language Proceedings of the 1969 on Computational linguistics, S\u00e5ng-S\u00e4by, Sweden, pages 1-3 Ce mod\u00e8le, partiellement influenc\u00e9 par les travaux de Sydney Lamb, a \u00e9t\u00e9 largement utilis\u00e9 par les \u00e9tudiants de Schank \u00e0 l'universit\u00e9 de Yale, tels que Robert Wilensky, Wendy Lehnert et Janet Kolodner.", "token2charspan": [[0, 5], [6, 12], [12, 13], [14, 18], [18, 19], [20, 21], [22, 32], [33, 43], [44, 50], [51, 54], [55, 62], [63, 71], [72, 83], [84, 86], [87, 90], [91, 95], [96, 98], [99, 112], [113, 124], [124, 125], [126, 135], [135, 136], [137, 143], [143, 144], [145, 150], [151, 154], [155, 157], [158, 164], [164, 165], [166, 179], [180, 189], [190, 193], [194, 197], [198, 205], [206, 208], [209, 215], [216, 220], [220, 221], [222, 223], [224, 227], [228, 237], [238, 245], [246, 249], [250, 253], [254, 263], [264, 266], [267, 273], [274, 275], [276, 278], [278, 288], [289, 291], [292, 296], [296, 297], [298, 302], [303, 306], [307, 313], [314, 322], [322, 323], [324, 329], [330, 337], [338, 340], [341, 346], [347, 355], [355, 356]]}
{"doc_key": "ai-dev-253", "ner": [[0, 5, "algorithm"], [8, 8, "algorithm"], [16, 16, "algorithm"], [18, 20, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[8, 8, 0, 5, "named", "", false, false], [16, 16, 0, 5, "named", "", false, false], [18, 20, 0, 5, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["La", "m\u00e9thode", "du", "maximum", "de", "vraisemblance", "am\u00e9lior\u00e9e", "(", "IMLM", ")", "est", "une", "combinaison", "de", "deux", "estimateurs", "MLM", "(", "maximum", "de", "vraisemblance", ")", "."], "sentence-detokenized": "La m\u00e9thode du maximum de vraisemblance am\u00e9lior\u00e9e (IMLM) est une combinaison de deux estimateurs MLM (maximum de vraisemblance).", "token2charspan": [[0, 2], [3, 10], [11, 13], [14, 21], [22, 24], [25, 38], [39, 48], [49, 50], [50, 54], [54, 55], [56, 59], [60, 63], [64, 75], [76, 78], [79, 83], [84, 95], [96, 99], [100, 101], [101, 108], [109, 111], [112, 125], [125, 126], [126, 127]]}
{"doc_key": "ai-dev-254", "ner": [[21, 23, "metrics"], [26, 28, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[26, 28, 21, 23, "named", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Ces", "m\u00e9thodes", "peuvent", "\u00e9galement", "analyser", "les", "r\u00e9sultats", "d'", "un", "programme", "et", "son", "utilit\u00e9", "et", "peuvent", "donc", "impliquer", "l'", "analyse", "de", "sa", "matrice", "de", "confusion", "(", "ou", "table", "de", "confusion", ")", "."], "sentence-detokenized": "Ces m\u00e9thodes peuvent \u00e9galement analyser les r\u00e9sultats d'un programme et son utilit\u00e9 et peuvent donc impliquer l'analyse de sa matrice de confusion (ou table de confusion).", "token2charspan": [[0, 3], [4, 12], [13, 20], [21, 30], [31, 39], [40, 43], [44, 53], [54, 56], [56, 58], [59, 68], [69, 71], [72, 75], [76, 83], [84, 86], [87, 94], [95, 99], [100, 109], [110, 112], [112, 119], [120, 122], [123, 125], [126, 133], [134, 136], [137, 146], [147, 148], [148, 150], [151, 156], [157, 159], [160, 169], [169, 170], [170, 171]]}
{"doc_key": "ai-dev-255", "ner": [[0, 0, "product"], [9, 10, "researcher"], [12, 13, "researcher"], [15, 17, "researcher"], [23, 31, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[0, 0, 9, 10, "origin", "", false, false], [0, 0, 12, 13, "origin", "", false, false], [0, 0, 15, 17, "origin", "", false, false], [0, 0, 23, 31, "temporal", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["SURF", "a", "\u00e9t\u00e9", "publi\u00e9", "pour", "la", "premi\u00e8re", "fois", "par", "Herbert", "Bay", ",", "Tinne", "Tuytelaars", "et", "Luc", "Van", "Gool", ",", "et", "pr\u00e9sent\u00e9", "\u00e0", "la", "Conf\u00e9rence", "europ\u00e9enne", "sur", "la", "vision", "par", "ordinateur", "de", "2006", "."], "sentence-detokenized": "SURF a \u00e9t\u00e9 publi\u00e9 pour la premi\u00e8re fois par Herbert Bay, Tinne Tuytelaars et Luc Van Gool, et pr\u00e9sent\u00e9 \u00e0 la Conf\u00e9rence europ\u00e9enne sur la vision par ordinateur de 2006.", "token2charspan": [[0, 4], [5, 6], [7, 10], [11, 17], [18, 22], [23, 25], [26, 34], [35, 39], [40, 43], [44, 51], [52, 55], [55, 56], [57, 62], [63, 73], [74, 76], [77, 80], [81, 84], [85, 89], [89, 90], [91, 93], [94, 102], [103, 104], [105, 107], [108, 118], [119, 129], [130, 133], [134, 136], [137, 143], [144, 147], [148, 158], [159, 161], [162, 166], [166, 167]]}
{"doc_key": "ai-dev-256", "ner": [[0, 3, "task"], [8, 10, "field"], [13, 14, "field"], [17, 19, "field"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[0, 3, 8, 10, "part-of", "", false, false], [0, 3, 13, 14, "part-of", "", false, false], [0, 3, 17, 19, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["L'", "OCR", "est", "un", "domaine", "de", "recherche", "en", "reconnaissance", "des", "formes", ",", "en", "intelligence", "artificielle", "et", "en", "vision", "par", "ordinateur", "."], "sentence-detokenized": "L'OCR est un domaine de recherche en reconnaissance des formes, en intelligence artificielle et en vision par ordinateur.", "token2charspan": [[0, 2], [2, 5], [6, 9], [10, 12], [13, 20], [21, 23], [24, 33], [34, 36], [37, 51], [52, 55], [56, 62], [62, 63], [64, 66], [67, 79], [80, 92], [93, 95], [96, 98], [99, 105], [106, 109], [110, 120], [120, 121]]}
{"doc_key": "ai-dev-257", "ner": [[9, 13, "metrics"], [16, 20, "algorithm"], [22, 22, "algorithm"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[22, 22, 16, 20, "named", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Si", "l'", "on", "poursuit", "l'", "exemple", "en", "utilisant", "l'", "estimateur", "du", "maximum", "de", "vraisemblance", ",", "la", "fonction", "de", "densit\u00e9", "de", "probabilit\u00e9", "(", "pdf", ")", "du", "bruit", "pour", "un", "\u00e9chantillon", "mathwn", "/", "math", "est", "la", "suivante"], "sentence-detokenized": "Si l'on poursuit l'exemple en utilisant l'estimateur du maximum de vraisemblance, la fonction de densit\u00e9 de probabilit\u00e9 (pdf) du bruit pour un \u00e9chantillon mathwn / math est la suivante", "token2charspan": [[0, 2], [3, 5], [5, 7], [8, 16], [17, 19], [19, 26], [27, 29], [30, 39], [40, 42], [42, 52], [53, 55], [56, 63], [64, 66], [67, 80], [80, 81], [82, 84], [85, 93], [94, 96], [97, 104], [105, 107], [108, 119], [120, 121], [121, 124], [124, 125], [126, 128], [129, 134], [135, 139], [140, 142], [143, 154], [155, 161], [162, 163], [164, 168], [169, 172], [173, 175], [176, 184]]}
{"doc_key": "ai-dev-258", "ner": [[4, 6, "field"], [9, 11, "task"], [14, 16, "task"], [19, 20, "task"], [23, 25, "task"], [28, 32, "task"], [35, 35, "task"], [38, 38, "task"], [41, 43, "task"], [46, 47, "task"], [50, 54, "task"], [57, 59, "task"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], "relations": [[9, 11, 4, 6, "part-of", "", false, false], [14, 16, 4, 6, "part-of", "", false, false], [19, 20, 4, 6, "part-of", "", false, false], [23, 25, 4, 6, "part-of", "", false, false], [28, 32, 4, 6, "part-of", "", false, false], [35, 35, 4, 6, "part-of", "", false, false], [38, 38, 4, 6, "part-of", "", false, false], [41, 43, 4, 6, "part-of", "", false, false], [46, 47, 4, 6, "part-of", "", false, false], [50, 54, 4, 6, "part-of", "", false, false], [57, 59, 4, 6, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10], "sentence": ["Les", "sous-domaines", "de", "la", "vision", "par", "ordinateur", "comprennent", "la", "reconstruction", "de", "sc\u00e8nes", ",", "la", "d\u00e9tection", "d'", "\u00e9v\u00e9nements", ",", "le", "suivi", "vid\u00e9o", ",", "la", "reconnaissance", "d'", "objets", ",", "l'", "estimation", "de", "pose", "en", "3D", ",", "l'", "apprentissage", ",", "l'", "indexation", ",", "l'", "estimation", "de", "mouvement", ",", "l'", "asservissement", "visuel", ",", "la", "mod\u00e9lisation", "de", "sc\u00e8nes", "en", "3D", "et", "la", "restauration", "d'", "images", "."], "sentence-detokenized": "Les sous-domaines de la vision par ordinateur comprennent la reconstruction de sc\u00e8nes, la d\u00e9tection d'\u00e9v\u00e9nements, le suivi vid\u00e9o, la reconnaissance d'objets, l'estimation de pose en 3D, l'apprentissage, l'indexation, l'estimation de mouvement, l'asservissement visuel, la mod\u00e9lisation de sc\u00e8nes en 3D et la restauration d'images.", "token2charspan": [[0, 3], [4, 17], [18, 20], [21, 23], [24, 30], [31, 34], [35, 45], [46, 57], [58, 60], [61, 75], [76, 78], [79, 85], [85, 86], [87, 89], [90, 99], [100, 102], [102, 112], [112, 113], [114, 116], [117, 122], [123, 128], [128, 129], [130, 132], [133, 147], [148, 150], [150, 156], [156, 157], [158, 160], [160, 170], [171, 173], [174, 178], [179, 181], [182, 184], [184, 185], [186, 188], [188, 201], [201, 202], [203, 205], [205, 215], [215, 216], [217, 219], [219, 229], [230, 232], [233, 242], [242, 243], [244, 246], [246, 260], [261, 267], [267, 268], [269, 271], [272, 284], [285, 287], [288, 294], [295, 297], [298, 300], [301, 303], [304, 306], [307, 319], [320, 322], [322, 328], [328, 329]]}
{"doc_key": "ai-dev-259", "ner": [[6, 10, "conference"], [12, 12, "researcher"], [16, 17, "misc"], [21, 23, "conference"], [25, 25, "researcher"], [27, 27, "researcher"], [31, 33, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[6, 10, 21, 23, "named", "", false, false], [12, 12, 16, 17, "win-defeat", "", false, false], [12, 12, 31, 33, "related-to", "writes_about", true, false], [16, 17, 6, 10, "temporal", "", false, false], [25, 25, 16, 17, "win-defeat", "", false, true], [25, 25, 31, 33, "related-to", "writes_about", true, false], [27, 27, 16, 17, "win-defeat", "", false, true], [27, 27, 31, 33, "related-to", "writes_about", true, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "sentence": ["En", "2013", ",", "lors", "de", "l'", "International", "Conference", "on", "Computer", "Vision", ",", "Terzopoulos", "a", "re\u00e7u", "un", "prix", "Helmholtz", "pour", "son", "article", "ICCV", "de", "1987", "avec", "Kass", "et", "Witkin", "sur", "les", "mod\u00e8les", "de", "contour", "actifs", "."], "sentence-detokenized": "En 2013, lors de l'International Conference on Computer Vision, Terzopoulos a re\u00e7u un prix Helmholtz pour son article ICCV de 1987 avec Kass et Witkin sur les mod\u00e8les de contour actifs.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 13], [14, 16], [17, 19], [19, 32], [33, 43], [44, 46], [47, 55], [56, 62], [62, 63], [64, 75], [76, 77], [78, 82], [83, 85], [86, 90], [91, 100], [101, 105], [106, 109], [110, 117], [118, 122], [123, 125], [126, 130], [131, 135], [136, 140], [141, 143], [144, 150], [151, 154], [155, 158], [159, 166], [167, 169], [170, 177], [178, 184], [184, 185]]}
{"doc_key": "ai-dev-260", "ner": [[22, 23, "task"], [26, 29, "algorithm"], [32, 34, "algorithm"], [35, 35, "algorithm"], [38, 40, "algorithm"], [44, 45, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[22, 23, 26, 29, "usage", "", true, false], [22, 23, 32, 34, "usage", "", true, false], [22, 23, 35, 35, "usage", "", true, false], [22, 23, 38, 40, "usage", "", true, false], [22, 23, 44, 45, "usage", "", true, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Si", "la", "fonction", "de", "r\u00e9gularisation", "Il", "existe", "de", "nombreux", "algorithmes", "pour", "r\u00e9soudre", "ce", "type", "de", "probl\u00e8mes", ";", "les", "plus", "populaires", "pour", "la", "classification", "lin\u00e9aire", "sont", "la", "descente", "de", "gradient", "stochastique", ",", "la", "descente", "de", "gradient", "L-BFGS", ",", "la", "descente", "de", "coordonn\u00e9es", "et", "les", "m\u00e9thodes", "de", "Newton", "."], "sentence-detokenized": "Si la fonction de r\u00e9gularisation Il existe de nombreux algorithmes pour r\u00e9soudre ce type de probl\u00e8mes ; les plus populaires pour la classification lin\u00e9aire sont la descente de gradient stochastique, la descente de gradient L-BFGS, la descente de coordonn\u00e9es et les m\u00e9thodes de Newton.", "token2charspan": [[0, 2], [3, 5], [6, 14], [15, 17], [18, 32], [33, 35], [36, 42], [43, 45], [46, 54], [55, 66], [67, 71], [72, 80], [81, 83], [84, 88], [89, 91], [92, 101], [102, 103], [104, 107], [108, 112], [113, 123], [124, 128], [129, 131], [132, 146], [147, 155], [156, 160], [161, 163], [164, 172], [173, 175], [176, 184], [185, 197], [197, 198], [199, 201], [202, 210], [211, 213], [214, 222], [223, 229], [229, 230], [231, 233], [234, 242], [243, 245], [246, 257], [258, 260], [261, 264], [265, 273], [274, 276], [277, 283], [283, 284]]}
{"doc_key": "ai-dev-261", "ner": [[3, 6, "algorithm"], [8, 8, "algorithm"], [14, 15, "researcher"], [17, 18, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[3, 6, 14, 15, "origin", "", false, false], [8, 8, 3, 6, "named", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Les", "r\u00e9seaux", "de", "m\u00e9moire", "\u00e0", "long", "terme", "(", "LSTM", ")", "ont", "\u00e9t\u00e9", "invent\u00e9s", "par", "Sepp", "Hochreiter", "et", "J\u00fcrgen", "Schmidhuber", "en", "1997", "et", "ont", "\u00e9tabli", "des", "records", "de", "pr\u00e9cision", "dans", "de", "multiples", "domaines", "d'", "application", "."], "sentence-detokenized": "Les r\u00e9seaux de m\u00e9moire \u00e0 long terme (LSTM) ont \u00e9t\u00e9 invent\u00e9s par Sepp Hochreiter et J\u00fcrgen Schmidhuber en 1997 et ont \u00e9tabli des records de pr\u00e9cision dans de multiples domaines d'application.", "token2charspan": [[0, 3], [4, 11], [12, 14], [15, 22], [23, 24], [25, 29], [30, 35], [36, 37], [37, 41], [41, 42], [43, 46], [47, 50], [51, 59], [60, 63], [64, 68], [69, 79], [80, 82], [83, 89], [90, 101], [102, 104], [105, 109], [110, 112], [113, 116], [117, 123], [124, 127], [128, 135], [136, 138], [139, 148], [149, 153], [154, 156], [157, 166], [167, 175], [176, 178], [178, 189], [189, 190]]}
{"doc_key": "ai-dev-262", "ner": [[0, 0, "product"], [5, 7, "organisation"]], "ner_mapping_to_source": [0, 1], "relations": [[0, 0, 5, 7, "physical", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["TN", "a", "\u00e9t\u00e9", "d\u00e9velopp\u00e9", "au", "Massachusetts", "General", "Hospital", "et", "a", "\u00e9t\u00e9", "test\u00e9", "dans", "de", "multiples", "sc\u00e9narios", ",", "notamment", "l'", "extraction", "du", "statut", "tabagique", ",", "les", "ant\u00e9c\u00e9dents", "familiaux", "de", "maladie", "coronarienne", ",", "l'", "identification", "des", "patients", "souffrant", "de", "troubles", "du", "sommeil", ","], "sentence-detokenized": "TN a \u00e9t\u00e9 d\u00e9velopp\u00e9 au Massachusetts General Hospital et a \u00e9t\u00e9 test\u00e9 dans de multiples sc\u00e9narios, notamment l'extraction du statut tabagique, les ant\u00e9c\u00e9dents familiaux de maladie coronarienne, l'identification des patients souffrant de troubles du sommeil,", "token2charspan": [[0, 2], [3, 4], [5, 8], [9, 18], [19, 21], [22, 35], [36, 43], [44, 52], [53, 55], [56, 57], [58, 61], [62, 67], [68, 72], [73, 75], [76, 85], [86, 95], [95, 96], [97, 106], [107, 109], [109, 119], [120, 122], [123, 129], [130, 139], [139, 140], [141, 144], [145, 156], [157, 166], [167, 169], [170, 177], [178, 190], [190, 191], [192, 194], [194, 208], [209, 212], [213, 221], [222, 231], [232, 234], [235, 243], [244, 246], [247, 254], [254, 255]]}
{"doc_key": "ai-dev-263", "ner": [[3, 4, "researcher"], [10, 10, "product"], [19, 20, "organisation"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[3, 4, 10, 10, "role", "sells", false, false], [10, 10, 19, 20, "physical", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["En", "1960", ",", "Devol", "a", "personnellement", "vendu", "le", "premier", "robot", "Unimate", ",", "qui", "a", "\u00e9t\u00e9", "exp\u00e9di\u00e9", "en", "1961", "\u00e0", "General", "Motors", "."], "sentence-detokenized": "En 1960, Devol a personnellement vendu le premier robot Unimate, qui a \u00e9t\u00e9 exp\u00e9di\u00e9 en 1961 \u00e0 General Motors.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 14], [15, 16], [17, 32], [33, 38], [39, 41], [42, 49], [50, 55], [56, 63], [63, 64], [65, 68], [69, 70], [71, 74], [75, 82], [83, 85], [86, 90], [91, 92], [93, 100], [101, 107], [107, 108]]}
{"doc_key": "ai-dev-264", "ner": [[1, 5, "conference"], [15, 16, "location"], [18, 18, "location"], [21, 21, "country"], [34, 35, "organisation"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[1, 5, 15, 16, "physical", "", false, false], [15, 16, 18, 18, "physical", "", false, false], [18, 18, 21, 21, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["La", "Campus", "Party", "Europe", "s'", "est", "tenue", "du", "14", "au", "18", "avril", "2010", "\u00e0", "la", "Caja", "M\u00e1gica", "de", "Madrid", ",", "en", "Espagne", ",", "avec", "800", "participants", "de", "chacun", "des", "27", "\u00c9tats", "membres", "de", "l'", "Union", "europ\u00e9enne", "."], "sentence-detokenized": "La Campus Party Europe s'est tenue du 14 au 18 avril 2010 \u00e0 la Caja M\u00e1gica de Madrid, en Espagne, avec 800 participants de chacun des 27 \u00c9tats membres de l'Union europ\u00e9enne.", "token2charspan": [[0, 2], [3, 9], [10, 15], [16, 22], [23, 25], [25, 28], [29, 34], [35, 37], [38, 40], [41, 43], [44, 46], [47, 52], [53, 57], [58, 59], [60, 62], [63, 67], [68, 74], [75, 77], [78, 84], [84, 85], [86, 88], [89, 96], [96, 97], [98, 102], [103, 106], [107, 119], [120, 122], [123, 129], [130, 133], [134, 136], [137, 142], [143, 150], [151, 153], [154, 156], [156, 161], [162, 172], [172, 173]]}
{"doc_key": "ai-dev-265", "ner": [[7, 7, "organisation"], [10, 12, "organisation"], [19, 26, "misc"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[19, 26, 7, 7, "origin", "", false, false], [19, 26, 10, 12, "origin", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["En", "juillet", "2016", ",", "une", "collaboration", "entre", "DeepMind", "et", "le", "Moorfields", "Eye", "Hospital", "a", "\u00e9t\u00e9", "annonc\u00e9e", "pour", "d\u00e9velopper", "des", "applications", "d'", "IA", "pour", "les", "soins", "de", "sant\u00e9", "."], "sentence-detokenized": "En juillet 2016, une collaboration entre DeepMind et le Moorfields Eye Hospital a \u00e9t\u00e9 annonc\u00e9e pour d\u00e9velopper des applications d'IA pour les soins de sant\u00e9.", "token2charspan": [[0, 2], [3, 10], [11, 15], [15, 16], [17, 20], [21, 34], [35, 40], [41, 49], [50, 52], [53, 55], [56, 66], [67, 70], [71, 79], [80, 81], [82, 85], [86, 94], [95, 99], [100, 110], [111, 114], [115, 127], [128, 130], [130, 132], [133, 137], [138, 141], [142, 147], [148, 150], [151, 156], [156, 157]]}
{"doc_key": "ai-dev-266", "ner": [[6, 6, "misc"], [13, 15, "university"], [17, 17, "university"], [19, 20, "university"], [22, 23, "university"], [25, 25, "university"], [27, 27, "university"], [30, 33, "university"], [35, 36, "university"], [38, 39, "university"], [41, 41, "university"], [44, 46, "university"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], "relations": [[6, 6, 13, 15, "physical", "", false, false], [6, 6, 17, 17, "physical", "", false, false], [6, 6, 19, 20, "physical", "", false, false], [6, 6, 22, 23, "physical", "", false, false], [6, 6, 25, 25, "physical", "", false, false], [6, 6, 27, 27, "physical", "", false, false], [6, 6, 30, 33, "physical", "", false, false], [6, 6, 35, 36, "physical", "", false, false], [6, 6, 38, 39, "physical", "", false, false], [6, 6, 41, 41, "physical", "", false, false], [6, 6, 44, 46, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10], "sentence": ["Ils", "ont", "fini", "par", "attribuer", "onze", "PR2", "\u00e0", "diff\u00e9rentes", "institutions", ",", "dont", "l'", "Universit\u00e9", "de", "Fribourg", ",", "Bosch", ",", "Georgia", "Tech", ",", "KU", "Leuven", ",", "MIT", ",", "Stanford", ",", "l'", "Universit\u00e9", "technique", "de", "Munich", ",", "UC", "Berkeley", ",", "U", "Penn", ",", "USC", "et", "l'", "Universit\u00e9", "de", "Tokyo", "."], "sentence-detokenized": "Ils ont fini par attribuer onze PR2 \u00e0 diff\u00e9rentes institutions, dont l'Universit\u00e9 de Fribourg, Bosch, Georgia Tech, KU Leuven, MIT, Stanford, l'Universit\u00e9 technique de Munich, UC Berkeley, U Penn, USC et l'Universit\u00e9 de Tokyo.", "token2charspan": [[0, 3], [4, 7], [8, 12], [13, 16], [17, 26], [27, 31], [32, 35], [36, 37], [38, 49], [50, 62], [62, 63], [64, 68], [69, 71], [71, 81], [82, 84], [85, 93], [93, 94], [95, 100], [100, 101], [102, 109], [110, 114], [114, 115], [116, 118], [119, 125], [125, 126], [127, 130], [130, 131], [132, 140], [140, 141], [142, 144], [144, 154], [155, 164], [165, 167], [168, 174], [174, 175], [176, 178], [179, 187], [187, 188], [189, 190], [191, 195], [195, 196], [197, 200], [201, 203], [204, 206], [206, 216], [217, 219], [220, 225], [225, 226]]}
{"doc_key": "ai-dev-267", "ner": [[3, 3, "metrics"], [5, 5, "metrics"], [7, 7, "metrics"], [9, 9, "metrics"], [17, 19, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[3, 3, 17, 19, "part-of", "", false, false], [5, 5, 17, 19, "part-of", "", false, false], [7, 7, 17, 19, "part-of", "", false, false], [9, 9, 17, 19, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Les", "comptes", "de", "TP", ",", "TN", ",", "FP", "et", "FN", "sont", "g\u00e9n\u00e9ralement", "conserv\u00e9s", "dans", "un", "tableau", "appel\u00e9", "matrice", "de", "confusion", "."], "sentence-detokenized": "Les comptes de TP, TN, FP et FN sont g\u00e9n\u00e9ralement conserv\u00e9s dans un tableau appel\u00e9 matrice de confusion.", "token2charspan": [[0, 3], [4, 11], [12, 14], [15, 17], [17, 18], [19, 21], [21, 22], [23, 25], [26, 28], [29, 31], [32, 36], [37, 49], [50, 59], [60, 64], [65, 67], [68, 75], [76, 82], [83, 90], [91, 93], [94, 103], [103, 104]]}
{"doc_key": "ai-dev-268", "ner": [[9, 11, "metrics"], [14, 15, "metrics"], [18, 19, "metrics"], [22, 23, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [], "relations_mapping_to_source": [], "sentence": ["Comme", "ensemble", "de", "caract\u00e9ristiques", ",", "on", "utilise", "g\u00e9n\u00e9ralement", "le", "gain", "d'", "information", ",", "l'", "entropie", "crois\u00e9e", ",", "l'", "information", "mutuelle", "et", "l'", "odds", "ratio", "."], "sentence-detokenized": "Comme ensemble de caract\u00e9ristiques, on utilise g\u00e9n\u00e9ralement le gain d'information, l'entropie crois\u00e9e, l'information mutuelle et l'odds ratio.", "token2charspan": [[0, 5], [6, 14], [15, 17], [18, 34], [34, 35], [36, 38], [39, 46], [47, 59], [60, 62], [63, 67], [68, 70], [70, 81], [81, 82], [83, 85], [85, 93], [94, 101], [101, 102], [103, 105], [105, 116], [117, 125], [126, 128], [129, 131], [131, 135], [136, 141], [141, 142]]}
{"doc_key": "ai-dev-269", "ner": [[12, 14, "task"], [17, 19, "task"], [22, 22, "task"], [25, 25, "task"], [28, 30, "task"], [32, 32, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[32, 32, 28, 30, "named", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Elle", "a", "\u00e9t\u00e9", "appliqu\u00e9e", "avec", "succ\u00e8s", "\u00e0", "divers", "probl\u00e8mes", ",", "dont", "le", "contr\u00f4le", "des", "robots", ",", "la", "programmation", "des", "ascenseurs", ",", "les", "t\u00e9l\u00e9communications", ",", "les", "dames", "et", "le", "jeu", "de", "Go", "(", "AlphaGo", ")", "."], "sentence-detokenized": "Elle a \u00e9t\u00e9 appliqu\u00e9e avec succ\u00e8s \u00e0 divers probl\u00e8mes, dont le contr\u00f4le des robots, la programmation des ascenseurs, les t\u00e9l\u00e9communications, les dames et le jeu de Go (AlphaGo).", "token2charspan": [[0, 4], [5, 6], [7, 10], [11, 20], [21, 25], [26, 32], [33, 34], [35, 41], [42, 51], [51, 52], [53, 57], [58, 60], [61, 69], [70, 73], [74, 80], [80, 81], [82, 84], [85, 98], [99, 102], [103, 113], [113, 114], [115, 118], [119, 137], [137, 138], [139, 142], [143, 148], [149, 151], [152, 154], [155, 158], [159, 161], [162, 164], [165, 166], [166, 173], [173, 174], [174, 175]]}
{"doc_key": "ai-dev-270", "ner": [[11, 15, "misc"], [20, 23, "university"], [25, 25, "location"], [28, 28, "location"], [32, 36, "location"], [40, 44, "location"], [46, 46, "location"], [47, 49, "country"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "relations": [[11, 15, 20, 23, "physical", "", false, false], [20, 23, 25, 25, "physical", "", false, false], [25, 25, 28, 28, "physical", "", false, false], [32, 36, 40, 44, "physical", "", false, false], [40, 44, 46, 46, "physical", "", false, false], [46, 46, 47, 49, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["En", "2018", ",", "ann\u00e9e", "inaugurale", "de", "la", "mission", "8", ",", "le", "site", "am\u00e9ricain", "s'", "est", "d\u00e9roul\u00e9", "sur", "le", "campus", "du", "Georgia", "Institute", "of", "Technology", "\u00e0", "Atlanta", ",", "en", "G\u00e9orgie", ",", "et", "le", "site", "Asie", "/", "Pacifique", "s'", "est", "d\u00e9roul\u00e9", "au", "gymnase", "de", "l'", "universit\u00e9", "Beihang", "\u00e0", "P\u00e9kin", ",", "en", "Chine", "."], "sentence-detokenized": "En 2018, ann\u00e9e inaugurale de la mission 8, le site am\u00e9ricain s'est d\u00e9roul\u00e9 sur le campus du Georgia Institute of Technology \u00e0 Atlanta, en G\u00e9orgie, et le site Asie/Pacifique s'est d\u00e9roul\u00e9 au gymnase de l'universit\u00e9 Beihang \u00e0 P\u00e9kin, en Chine.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 14], [15, 25], [26, 28], [29, 31], [32, 39], [40, 41], [41, 42], [43, 45], [46, 50], [51, 60], [61, 63], [63, 66], [67, 74], [75, 78], [79, 81], [82, 88], [89, 91], [92, 99], [100, 109], [110, 112], [113, 123], [124, 125], [126, 133], [133, 134], [135, 137], [138, 145], [145, 146], [147, 149], [150, 152], [153, 157], [158, 162], [162, 163], [163, 172], [173, 175], [175, 178], [179, 186], [187, 189], [190, 197], [198, 200], [201, 203], [203, 213], [214, 221], [222, 223], [224, 229], [229, 230], [231, 233], [234, 239], [239, 240]]}
{"doc_key": "ai-dev-271", "ner": [[0, 2, "field"], [8, 10, "field"], [15, 16, "field"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[0, 2, 8, 10, "origin", "", false, false], [0, 2, 8, 10, "related-to", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["L'", "apprentissage", "automatique", "est", "fortement", "li\u00e9", "\u00e0", "la", "reconnaissance", "des", "formes", "et", "d\u00e9coule", "de", "l'", "intelligence", "artificielle", "."], "sentence-detokenized": "L'apprentissage automatique est fortement li\u00e9 \u00e0 la reconnaissance des formes et d\u00e9coule de l'intelligence artificielle.", "token2charspan": [[0, 2], [2, 15], [16, 27], [28, 31], [32, 41], [42, 45], [46, 47], [48, 50], [51, 65], [66, 69], [70, 76], [77, 79], [80, 87], [88, 90], [91, 93], [93, 105], [106, 118], [118, 119]]}
{"doc_key": "ai-dev-272", "ner": [[6, 6, "programlang"], [17, 18, "product"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "est", "livr\u00e9", "avec", "3", "jeux", "Java", "qui", "sont", "contr\u00f4l\u00e9s", "par", "la", "t\u00e9l\u00e9commande", "et", "affich\u00e9s", "sur", "son", "\u00e9cran", "LCD", "."], "sentence-detokenized": "Il est livr\u00e9 avec 3 jeux Java qui sont contr\u00f4l\u00e9s par la t\u00e9l\u00e9commande et affich\u00e9s sur son \u00e9cran LCD.", "token2charspan": [[0, 2], [3, 6], [7, 12], [13, 17], [18, 19], [20, 24], [25, 29], [30, 33], [34, 38], [39, 48], [49, 52], [53, 55], [56, 68], [69, 71], [72, 80], [81, 84], [85, 88], [89, 94], [95, 98], [98, 99]]}
{"doc_key": "ai-dev-273", "ner": [[9, 21, "task"], [1, 4, "algorithm"]], "ner_mapping_to_source": [0, 1], "relations": [[1, 4, 9, 21, "type-of", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["La", "capture", "de", "mouvement", "optique", "est", "une", "technique", "d'", "estimation", "de", "la", "pose", "des", "corps", "articul\u00e9s", "bas\u00e9e", "sur", "la", "vision", "par", "ordinateur", "qui", "conna\u00eet", "un", "grand", "succ\u00e8s", "commercial", "mais", "qui", "est", "sp\u00e9cialis\u00e9e", "."], "sentence-detokenized": "La capture de mouvement optique est une technique d'estimation de la pose des corps articul\u00e9s bas\u00e9e sur la vision par ordinateur qui conna\u00eet un grand succ\u00e8s commercial mais qui est sp\u00e9cialis\u00e9e.", "token2charspan": [[0, 2], [3, 10], [11, 13], [14, 23], [24, 31], [32, 35], [36, 39], [40, 49], [50, 52], [52, 62], [63, 65], [66, 68], [69, 73], [74, 77], [78, 83], [84, 93], [94, 99], [100, 103], [104, 106], [107, 113], [114, 117], [118, 128], [129, 132], [133, 140], [141, 143], [144, 149], [150, 156], [157, 167], [168, 172], [173, 176], [177, 180], [181, 192], [192, 193]]}
{"doc_key": "ai-dev-274", "ner": [[0, 1, "organisation"], [7, 8, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[0, 1, 7, 8, "compare", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Le", "SMC", "est", "tr\u00e8s", "similaire", "\u00e0", "l'", "indice", "Jaccard", ",", "plus", "populaire", "."], "sentence-detokenized": "Le SMC est tr\u00e8s similaire \u00e0 l'indice Jaccard, plus populaire.", "token2charspan": [[0, 2], [3, 6], [7, 10], [11, 15], [16, 25], [26, 27], [28, 30], [30, 36], [37, 44], [44, 45], [46, 50], [51, 60], [60, 61]]}
{"doc_key": "ai-dev-275", "ner": [[1, 1, "product"], [3, 7, "product"], [10, 13, "product"], [22, 23, "researcher"], [32, 32, "organisation"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[1, 1, 10, 13, "named", "", false, false], [1, 1, 22, 23, "artifact", "", false, false], [1, 1, 32, 32, "artifact", "", false, false], [3, 7, 1, 1, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Le", "PUMA", "(", "Programmable", "Universal", "Machine", "for", "Assembly", ",", "ou", "Programmable", "Universal", "Manipulation", "Arm", ")", "est", "un", "bras", "robotique", "industriel", "d\u00e9velopp\u00e9", "par", "Victor", "Scheinman", "au", "sein", "de", "la", "soci\u00e9t\u00e9", "pionni\u00e8re", "de", "robots", "Unimation", "."], "sentence-detokenized": "Le PUMA (Programmable Universal Machine for Assembly, ou Programmable Universal Manipulation Arm) est un bras robotique industriel d\u00e9velopp\u00e9 par Victor Scheinman au sein de la soci\u00e9t\u00e9 pionni\u00e8re de robots Unimation.", "token2charspan": [[0, 2], [3, 7], [8, 9], [9, 21], [22, 31], [32, 39], [40, 43], [44, 52], [52, 53], [54, 56], [57, 69], [70, 79], [80, 92], [93, 96], [96, 97], [98, 101], [102, 104], [105, 109], [110, 119], [120, 130], [131, 140], [141, 144], [145, 151], [152, 161], [162, 164], [165, 169], [170, 172], [173, 175], [176, 183], [184, 193], [194, 196], [197, 203], [204, 213], [213, 214]]}
{"doc_key": "ai-dev-276", "ner": [[4, 4, "programlang"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "est", "\u00e9crit", "en", "Python", "."], "sentence-detokenized": "Il est \u00e9crit en Python.", "token2charspan": [[0, 2], [3, 6], [7, 12], [13, 15], [16, 22], [22, 23]]}
{"doc_key": "ai-dev-277", "ner": [[0, 3, "misc"], [5, 5, "misc"], [17, 17, "field"], [20, 23, "field"], [26, 27, "field"], [33, 35, "field"], [38, 38, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 6, 7], "relations": [[0, 3, 5, 5, "related-to", "metric_for", true, false], [0, 3, 17, 17, "part-of", "", false, false], [0, 3, 20, 23, "part-of", "", false, false], [0, 3, 26, 27, "part-of", "", false, false], [0, 3, 33, 35, "part-of", "", false, false], [0, 3, 38, 38, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 5, 6], "sentence": ["La", "largeur", "de", "bande", "en", "hertz", "est", "un", "concept", "central", "dans", "de", "nombreux", "domaines", ",", "notamment", "l'", "\u00e9lectronique", ",", "la", "th\u00e9orie", "de", "l'", "information", ",", "les", "communications", "num\u00e9riques", ",", "les", "radiocommunications", ",", "le", "traitement", "du", "signal", "et", "la", "spectroscopie", ".", "Elle", "est", "l'", "un", "des", "facteurs", "d\u00e9terminants", "de", "la", "capacit\u00e9", "d'", "un", "canal", "de", "communication", "donn\u00e9", "."], "sentence-detokenized": "La largeur de bande en hertz est un concept central dans de nombreux domaines, notamment l'\u00e9lectronique, la th\u00e9orie de l'information, les communications num\u00e9riques, les radiocommunications, le traitement du signal et la spectroscopie. Elle est l'un des facteurs d\u00e9terminants de la capacit\u00e9 d'un canal de communication donn\u00e9.", "token2charspan": [[0, 2], [3, 10], [11, 13], [14, 19], [20, 22], [23, 28], [29, 32], [33, 35], [36, 43], [44, 51], [52, 56], [57, 59], [60, 68], [69, 77], [77, 78], [79, 88], [89, 91], [91, 103], [103, 104], [105, 107], [108, 115], [116, 118], [119, 121], [121, 132], [132, 133], [134, 137], [138, 152], [153, 163], [163, 164], [165, 168], [169, 188], [188, 189], [190, 192], [193, 203], [204, 206], [207, 213], [214, 216], [217, 219], [220, 233], [233, 234], [235, 239], [240, 243], [244, 246], [246, 248], [249, 252], [253, 261], [262, 274], [275, 277], [278, 280], [281, 289], [290, 292], [292, 294], [295, 300], [301, 303], [304, 317], [318, 323], [323, 324]]}
{"doc_key": "ai-dev-278", "ner": [[9, 9, "algorithm"], [11, 11, "algorithm"], [18, 21, "misc"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[9, 9, 18, 21, "part-of", "", false, false], [11, 11, 18, 21, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Si", "une", "perte", "convexe", "est", "utilis\u00e9e", "(", "comme", "dans", "AdaBoost", ",", "LogitBoost", "et", "tous", "les", "membres", "de", "la", "famille", "d'", "algorithmes", "AnyBoost", ")", ",", "un", "exemple", "avec", "une", "marge", "plus", "\u00e9lev\u00e9e", "recevra", "moins", "de", "poids", "(", "ou", "un", "poids", "\u00e9gal", ")", "qu'", "un", "exemple", "avec", "une", "marge", "plus", "faible", "."], "sentence-detokenized": "Si une perte convexe est utilis\u00e9e (comme dans AdaBoost, LogitBoost et tous les membres de la famille d'algorithmes AnyBoost), un exemple avec une marge plus \u00e9lev\u00e9e recevra moins de poids (ou un poids \u00e9gal) qu'un exemple avec une marge plus faible.", "token2charspan": [[0, 2], [3, 6], [7, 12], [13, 20], [21, 24], [25, 33], [34, 35], [35, 40], [41, 45], [46, 54], [54, 55], [56, 66], [67, 69], [70, 74], [75, 78], [79, 86], [87, 89], [90, 92], [93, 100], [101, 103], [103, 114], [115, 123], [123, 124], [124, 125], [126, 128], [129, 136], [137, 141], [142, 145], [146, 151], [152, 156], [157, 163], [164, 171], [172, 177], [178, 180], [181, 186], [187, 188], [188, 190], [191, 193], [194, 199], [200, 204], [204, 205], [206, 209], [209, 211], [212, 219], [220, 224], [225, 228], [229, 234], [235, 239], [240, 246], [246, 247]]}
{"doc_key": "ai-dev-279", "ner": [[5, 6, "researcher"], [9, 10, "researcher"]], "ner_mapping_to_source": [0, 1], "relations": [[5, 6, 9, 10, "named", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Le", "m\u00e9moire", "de", "dipl\u00f4me", "de", "Sepp", "Hochreiter", "de", "1991", "Sepp", "Hochreiter", "."], "sentence-detokenized": "Le m\u00e9moire de dipl\u00f4me de Sepp Hochreiter de 1991 Sepp Hochreiter.", "token2charspan": [[0, 2], [3, 10], [11, 13], [14, 21], [22, 24], [25, 29], [30, 40], [41, 43], [44, 48], [49, 53], [54, 64], [64, 65]]}
{"doc_key": "ai-dev-280", "ner": [[6, 7, "algorithm"], [9, 9, "algorithm"], [14, 17, "algorithm"], [19, 19, "algorithm"], [23, 25, "algorithm"], [27, 27, "algorithm"], [33, 35, "algorithm"], [39, 41, "algorithm"], [44, 45, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "relations": [[9, 9, 6, 7, "named", "", false, false], [19, 19, 14, 17, "named", "", false, false], [23, 25, 33, 35, "related-to", "", true, false], [27, 27, 23, 25, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Les", "mod\u00e8les", "discriminants", "typiques", "comprennent", "la", "r\u00e9gression", "logistique", "(", "LR", ")", ",", "les", "machines", "\u00e0", "vecteurs", "de", "support", "(", "SVM", ")", ",", "les", "champs", "al\u00e9atoires", "conditionnels", "(", "CRF", ")", "(", "sp\u00e9cifi\u00e9s", "sur", "un", "graphe", "non", "orient\u00e9", ")", ",", "les", "arbres", "de", "d\u00e9cision", ",", "les", "r\u00e9seaux", "neuronaux", "et", "bien", "d'", "autres", "."], "sentence-detokenized": "Les mod\u00e8les discriminants typiques comprennent la r\u00e9gression logistique (LR), les machines \u00e0 vecteurs de support (SVM), les champs al\u00e9atoires conditionnels (CRF) (sp\u00e9cifi\u00e9s sur un graphe non orient\u00e9), les arbres de d\u00e9cision, les r\u00e9seaux neuronaux et bien d'autres.", "token2charspan": [[0, 3], [4, 11], [12, 25], [26, 34], [35, 46], [47, 49], [50, 60], [61, 71], [72, 73], [73, 75], [75, 76], [76, 77], [78, 81], [82, 90], [91, 92], [93, 101], [102, 104], [105, 112], [113, 114], [114, 117], [117, 118], [118, 119], [120, 123], [124, 130], [131, 141], [142, 155], [156, 157], [157, 160], [160, 161], [162, 163], [163, 172], [173, 176], [177, 179], [180, 186], [187, 190], [191, 198], [198, 199], [199, 200], [201, 204], [205, 211], [212, 214], [215, 223], [223, 224], [225, 228], [229, 236], [237, 246], [247, 249], [250, 254], [255, 257], [257, 263], [263, 264]]}
{"doc_key": "ai-dev-281", "ner": [[12, 14, "metrics"], [36, 38, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "est", "\u00e9galement", "possible", "d'", "utiliser", "ces", "probabilit\u00e9s", "et", "d'", "\u00e9valuer", "l'", "erreur", "quadratique", "moyenne", "(", "ou", "une", "autre", "mesure", "similaire", ")", "entre", "les", "probabilit\u00e9s", "et", "les", "valeurs", "r\u00e9elles", ",", "puis", "de", "la", "combiner", "avec", "la", "matrice", "de", "confusion", "pour", "cr\u00e9er", "des", "fonctions", "de", "fitness", "tr\u00e8s", "efficaces", "pour", "la", "r\u00e9gression", "logistique", "."], "sentence-detokenized": "Il est \u00e9galement possible d'utiliser ces probabilit\u00e9s et d'\u00e9valuer l'erreur quadratique moyenne (ou une autre mesure similaire) entre les probabilit\u00e9s et les valeurs r\u00e9elles, puis de la combiner avec la matrice de confusion pour cr\u00e9er des fonctions de fitness tr\u00e8s efficaces pour la r\u00e9gression logistique.", "token2charspan": [[0, 2], [3, 6], [7, 16], [17, 25], [26, 28], [28, 36], [37, 40], [41, 53], [54, 56], [57, 59], [59, 66], [67, 69], [69, 75], [76, 87], [88, 95], [96, 97], [97, 99], [100, 103], [104, 109], [110, 116], [117, 126], [126, 127], [128, 133], [134, 137], [138, 150], [151, 153], [154, 157], [158, 165], [166, 173], [173, 174], [175, 179], [180, 182], [183, 185], [186, 194], [195, 199], [200, 202], [203, 210], [211, 213], [214, 223], [224, 228], [229, 234], [235, 238], [239, 248], [249, 251], [252, 259], [260, 264], [265, 274], [275, 279], [280, 282], [283, 293], [294, 304], [304, 305]]}
{"doc_key": "ai-dev-282", "ner": [[0, 2, "product"], [10, 13, "product"]], "ner_mapping_to_source": [0, 1], "relations": [[0, 2, 10, 13, "part-of", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["VoiceOver", "est", "apparu", "pour", "la", "premi\u00e8re", "fois", "en", "2005", "dans", "Mac", "OS", "X", "Tiger", "(", "10.4", ")", "."], "sentence-detokenized": "VoiceOver est apparu pour la premi\u00e8re fois en 2005 dans Mac OS X Tiger (10.4).", "token2charspan": [[0, 9], [10, 13], [14, 20], [21, 25], [26, 28], [29, 37], [38, 42], [43, 45], [46, 50], [51, 55], [56, 59], [60, 62], [63, 64], [65, 70], [71, 72], [72, 76], [76, 77], [77, 78]]}
{"doc_key": "ai-dev-283", "ner": [[15, 16, "algorithm"], [19, 21, "misc"], [26, 27, "metrics"], [30, 32, "algorithm"], [65, 68, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[15, 16, 19, 21, "related-to", "applied_to", false, false], [26, 27, 19, 21, "type-of", "", false, false], [26, 27, 30, 32, "related-to", "used_for", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["En", "pratique", ",", "les", "algorithmes", "d'", "apprentissage", "automatique", "y", "font", "face", "soit", "en", "employant", "une", "approximation", "convexe", "de", "la", "fonction", "de", "perte", "0-1", "(", "comme", "la", "perte", "charni\u00e8re", "pour", "le", "Support", "vector", "machine", ")", ",", "qui", "est", "plus", "facile", "\u00e0", "optimiser", ",", "soit", "en", "imposant", "des", "hypoth\u00e8ses", "sur", "la", "distribution", "mathP", "(", "x", ",", "y", ")", "/", "math", "(", "et", "cessent", "donc", "d'", "\u00eatre", "des", "algorithmes", "d'", "apprentissage", "agnostiques", "auxquels", "le", "r\u00e9sultat", "ci-dessus", "s'", "applique", ")", "."], "sentence-detokenized": "En pratique, les algorithmes d'apprentissage automatique y font face soit en employant une approximation convexe de la fonction de perte 0-1 (comme la perte charni\u00e8re pour le Support vector machine), qui est plus facile \u00e0 optimiser, soit en imposant des hypoth\u00e8ses sur la distribution mathP (x, y) / math (et cessent donc d'\u00eatre des algorithmes d'apprentissage agnostiques auxquels le r\u00e9sultat ci-dessus s'applique).", "token2charspan": [[0, 2], [3, 11], [11, 12], [13, 16], [17, 28], [29, 31], [31, 44], [45, 56], [57, 58], [59, 63], [64, 68], [69, 73], [74, 76], [77, 86], [87, 90], [91, 104], [105, 112], [113, 115], [116, 118], [119, 127], [128, 130], [131, 136], [137, 140], [141, 142], [142, 147], [148, 150], [151, 156], [157, 166], [167, 171], [172, 174], [175, 182], [183, 189], [190, 197], [197, 198], [198, 199], [200, 203], [204, 207], [208, 212], [213, 219], [220, 221], [222, 231], [231, 232], [233, 237], [238, 240], [241, 249], [250, 253], [254, 264], [265, 268], [269, 271], [272, 284], [285, 290], [291, 292], [292, 293], [293, 294], [295, 296], [296, 297], [298, 299], [300, 304], [305, 306], [306, 308], [309, 316], [317, 321], [322, 324], [324, 328], [329, 332], [333, 344], [345, 347], [347, 360], [361, 372], [373, 381], [382, 384], [385, 393], [394, 403], [404, 406], [406, 414], [414, 415], [415, 416]]}
{"doc_key": "ai-dev-284", "ner": [[0, 0, "misc"], [13, 16, "field"], [25, 25, "product"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[0, 0, 13, 16, "usage", "", false, false], [0, 0, 25, 25, "topic", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Westworld", "(", "1973", ")", "a", "\u00e9t\u00e9", "le", "premier", "long", "m\u00e9trage", "\u00e0", "utiliser", "le", "traitement", "num\u00e9rique", "des", "images", "pour", "simuler", "le", "point", "de", "vue", "d'", "un", "andro\u00efde", "."], "sentence-detokenized": "Westworld (1973) a \u00e9t\u00e9 le premier long m\u00e9trage \u00e0 utiliser le traitement num\u00e9rique des images pour simuler le point de vue d'un andro\u00efde.", "token2charspan": [[0, 9], [10, 11], [11, 15], [15, 16], [17, 18], [19, 22], [23, 25], [26, 33], [34, 38], [39, 46], [47, 48], [49, 57], [58, 60], [61, 71], [72, 81], [82, 85], [86, 92], [93, 97], [98, 105], [106, 108], [109, 114], [115, 117], [118, 121], [122, 124], [124, 126], [127, 135], [135, 136]]}
{"doc_key": "ai-dev-285", "ner": [[8, 11, "task"], [14, 15, "task"], [18, 18, "task"], [20, 21, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [], "relations_mapping_to_source": [], "sentence": ["Elle", "est", "maintenant", "aussi", "couramment", "utilis\u00e9e", "dans", "la", "reconnaissance", "de", "la", "parole", ",", "la", "synth\u00e8se", "vocale", ",", "la", "diarisation", ",", "Xavier", "Anguera", "et", "al", "."], "sentence-detokenized": "Elle est maintenant aussi couramment utilis\u00e9e dans la reconnaissance de la parole, la synth\u00e8se vocale, la diarisation, Xavier Anguera et al.", "token2charspan": [[0, 4], [5, 8], [9, 19], [20, 25], [26, 36], [37, 45], [46, 50], [51, 53], [54, 68], [69, 71], [72, 74], [75, 81], [81, 82], [83, 85], [86, 94], [95, 101], [101, 102], [103, 105], [106, 117], [117, 118], [119, 125], [126, 133], [134, 136], [137, 139], [139, 140]]}
{"doc_key": "ai-dev-286", "ner": [[9, 13, "algorithm"], [18, 19, "algorithm"], [22, 24, "algorithm"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[18, 19, 9, 13, "type-of", "", false, false], [22, 24, 9, 13, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Ici", ",", "math", "\\", "sigma", "/", "math", "est", "une", "fonction", "d'", "activation", "par", "\u00e9l\u00e9ments", ",", "telle", "qu'", "une", "fonction", "sigmo\u00efde", "ou", "une", "unit\u00e9", "lin\u00e9aire", "rectifi\u00e9e", "."], "sentence-detokenized": "Ici, math\\ sigma / math est une fonction d'activation par \u00e9l\u00e9ments, telle qu'une fonction sigmo\u00efde ou une unit\u00e9 lin\u00e9aire rectifi\u00e9e.", "token2charspan": [[0, 3], [3, 4], [5, 9], [9, 10], [11, 16], [17, 18], [19, 23], [24, 27], [28, 31], [32, 40], [41, 43], [43, 53], [54, 57], [58, 66], [66, 67], [68, 73], [74, 77], [77, 80], [81, 89], [90, 98], [99, 101], [102, 105], [106, 111], [112, 120], [121, 130], [130, 131]]}
{"doc_key": "ai-dev-287", "ner": [[11, 18, "algorithm"], [31, 31, "misc"], [35, 35, "misc"], [29, 39, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [], "relations_mapping_to_source": [], "sentence": ["Les", "approches", "traditionnelles", "bas\u00e9es", "sur", "la", "phon\u00e9tique", "(", "c'est-\u00e0-dire", "tous", "les", "mod\u00e8les", "bas\u00e9s", "sur", "le", "mod\u00e8le", "de", "Markov", "cach\u00e9", ")", "n\u00e9cessitaient", "des", "composants", "et", "un", "entra\u00eenement", "s\u00e9par\u00e9s", "pour", "le", "mod\u00e8le", "de", "prononciation", ",", "le", "mod\u00e8le", "acoustique", "et", "le", "mod\u00e8le", "linguistique", "."], "sentence-detokenized": "Les approches traditionnelles bas\u00e9es sur la phon\u00e9tique (c'est-\u00e0-dire tous les mod\u00e8les bas\u00e9s sur le mod\u00e8le de Markov cach\u00e9) n\u00e9cessitaient des composants et un entra\u00eenement s\u00e9par\u00e9s pour le mod\u00e8le de prononciation, le mod\u00e8le acoustique et le mod\u00e8le linguistique.", "token2charspan": [[0, 3], [4, 13], [14, 29], [30, 36], [37, 40], [41, 43], [44, 54], [55, 56], [56, 68], [69, 73], [74, 77], [78, 85], [86, 91], [92, 95], [96, 98], [99, 105], [106, 108], [109, 115], [116, 121], [121, 122], [123, 136], [137, 140], [141, 151], [152, 154], [155, 157], [158, 170], [171, 178], [179, 183], [184, 186], [187, 193], [194, 196], [197, 210], [210, 211], [212, 214], [215, 221], [222, 232], [233, 235], [236, 238], [239, 245], [246, 258], [258, 259]]}
{"doc_key": "ai-dev-288", "ner": [[1, 4, "algorithm"], [9, 11, "field"], [14, 16, "field"], [19, 21, "task"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[1, 4, 19, 21, "related-to", "used_for", false, false], [9, 11, 1, 4, "usage", "", false, false], [14, 16, 1, 4, "usage", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["L'", "op\u00e9rateur", "croix", "de", "Roberts", "est", "utilis\u00e9", "dans", "le", "traitement", "des", "images", "et", "la", "vision", "par", "ordinateur", "pour", "la", "d\u00e9tection", "des", "bords", "."], "sentence-detokenized": "L'op\u00e9rateur croix de Roberts est utilis\u00e9 dans le traitement des images et la vision par ordinateur pour la d\u00e9tection des bords.", "token2charspan": [[0, 2], [2, 11], [12, 17], [18, 20], [21, 28], [29, 32], [33, 40], [41, 45], [46, 48], [49, 59], [60, 63], [64, 70], [71, 73], [74, 76], [77, 83], [84, 87], [88, 98], [99, 103], [104, 106], [107, 116], [117, 120], [121, 126], [126, 127]]}
{"doc_key": "ai-dev-289", "ner": [[3, 3, "metrics"], [6, 6, "metrics"], [22, 22, "metrics"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[3, 3, 22, 22, "opposite", "", false, false], [6, 6, 22, 22, "opposite", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Les", "valeurs", "de", "sensibilit\u00e9", "et", "de", "sp\u00e9cificit\u00e9", "sont", "ind\u00e9pendantes", "du", "pourcentage", "de", "cas", "positifs", "dans", "la", "population", "concern\u00e9e", "(", "contrairement", "\u00e0", "la", "pr\u00e9cision", ",", "par", "exemple", ")", "."], "sentence-detokenized": "Les valeurs de sensibilit\u00e9 et de sp\u00e9cificit\u00e9 sont ind\u00e9pendantes du pourcentage de cas positifs dans la population concern\u00e9e (contrairement \u00e0 la pr\u00e9cision, par exemple).", "token2charspan": [[0, 3], [4, 11], [12, 14], [15, 26], [27, 29], [30, 32], [33, 44], [45, 49], [50, 63], [64, 66], [67, 78], [79, 81], [82, 85], [86, 94], [95, 99], [100, 102], [103, 113], [114, 123], [124, 125], [125, 138], [139, 140], [141, 143], [144, 153], [153, 154], [155, 158], [159, 166], [166, 167], [167, 168]]}
{"doc_key": "ai-dev-290", "ner": [[2, 3, "algorithm"], [12, 12, "misc"], [14, 15, "researcher"], [17, 18, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[12, 12, 2, 3, "topic", "", false, false], [12, 12, 14, 15, "artifact", "", false, false], [12, 12, 17, 18, "artifact", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Mais", "les", "mod\u00e8les", "perceptron", "ont", "\u00e9t\u00e9", "rendus", "tr\u00e8s", "impopulaires", "par", "le", "livre", "Perceptrons", "de", "Marvin", "Minsky", "et", "Seymour", "Papert", ",", "publi\u00e9", "en", "1969", "."], "sentence-detokenized": "Mais les mod\u00e8les perceptron ont \u00e9t\u00e9 rendus tr\u00e8s impopulaires par le livre Perceptrons de Marvin Minsky et Seymour Papert, publi\u00e9 en 1969.", "token2charspan": [[0, 4], [5, 8], [9, 16], [17, 27], [28, 31], [32, 35], [36, 42], [43, 47], [48, 60], [61, 64], [65, 67], [68, 73], [74, 85], [86, 88], [89, 95], [96, 102], [103, 105], [106, 113], [114, 120], [120, 121], [122, 128], [129, 131], [132, 136], [136, 137]]}
{"doc_key": "ai-dev-291", "ner": [[1, 6, "conference"], [13, 13, "organisation"], [29, 30, "task"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[1, 6, 29, 30, "topic", "", false, false], [13, 13, 1, 6, "role", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Les", "conf\u00e9rences", "sur", "la", "compr\u00e9hension", "des", "documents", ",", "organis\u00e9es", "chaque", "ann\u00e9e", "par", "le", "NIST", ",", "ont", "d\u00e9velopp\u00e9", "des", "crit\u00e8res", "d'", "\u00e9valuation", "sophistiqu\u00e9s", "pour", "les", "techniques", "acceptant", "le", "d\u00e9fi", "du", "r\u00e9sum\u00e9", "multi-document", "."], "sentence-detokenized": "Les conf\u00e9rences sur la compr\u00e9hension des documents, organis\u00e9es chaque ann\u00e9e par le NIST, ont d\u00e9velopp\u00e9 des crit\u00e8res d'\u00e9valuation sophistiqu\u00e9s pour les techniques acceptant le d\u00e9fi du r\u00e9sum\u00e9 multi-document.", "token2charspan": [[0, 3], [4, 15], [16, 19], [20, 22], [23, 36], [37, 40], [41, 50], [50, 51], [52, 62], [63, 69], [70, 75], [76, 79], [80, 82], [83, 87], [87, 88], [89, 92], [93, 102], [103, 106], [107, 115], [116, 118], [118, 128], [129, 141], [142, 146], [147, 150], [151, 161], [162, 171], [172, 174], [175, 179], [180, 182], [183, 189], [190, 204], [204, 205]]}
{"doc_key": "ai-dev-292", "ner": [[1, 2, "product"], [30, 32, "product"]], "ner_mapping_to_source": [0, 1], "relations": [[1, 2, 30, 32, "compare", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Un", "manipulateur", "parall\u00e8le", "est", "con\u00e7u", "de", "telle", "sorte", "que", "chaque", "cha\u00eene", "est", "g\u00e9n\u00e9ralement", "courte", ",", "simple", "et", "peut", "donc", "\u00eatre", "rigide", "contre", "tout", "mouvement", "ind\u00e9sirable", ",", "par", "rapport", "\u00e0", "un", "manipulateur", "en", "s\u00e9rie", "."], "sentence-detokenized": "Un manipulateur parall\u00e8le est con\u00e7u de telle sorte que chaque cha\u00eene est g\u00e9n\u00e9ralement courte, simple et peut donc \u00eatre rigide contre tout mouvement ind\u00e9sirable, par rapport \u00e0 un manipulateur en s\u00e9rie.", "token2charspan": [[0, 2], [3, 15], [16, 25], [26, 29], [30, 35], [36, 38], [39, 44], [45, 50], [51, 54], [55, 61], [62, 68], [69, 72], [73, 85], [86, 92], [92, 93], [94, 100], [101, 103], [104, 108], [109, 113], [114, 118], [119, 125], [126, 132], [133, 137], [138, 147], [148, 159], [159, 160], [161, 164], [165, 172], [173, 174], [175, 177], [178, 190], [191, 193], [194, 199], [199, 200]]}
{"doc_key": "ai-dev-293", "ner": [[29, 29, "misc"], [32, 35, "misc"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Le", "manipulateur", "est", "ce", "qui", "permet", "au", "robot", "de", "se", "d\u00e9placer", ",", "et", "la", "conception", "de", "ces", "syst\u00e8mes", "peut", "\u00eatre", "class\u00e9e", "en", "plusieurs", "types", "courants", ",", "tels", "que", "le", "SCARA", "et", "le", "robot", "\u00e0", "coordonn\u00e9es", "cart\u00e9siennes", ",", "qui", "utilisent", "diff\u00e9rents", "syst\u00e8mes", "de", "coordonn\u00e9es", "pour", "diriger", "les", "bras", "de", "la", "machine", "."], "sentence-detokenized": "Le manipulateur est ce qui permet au robot de se d\u00e9placer, et la conception de ces syst\u00e8mes peut \u00eatre class\u00e9e en plusieurs types courants, tels que le SCARA et le robot \u00e0 coordonn\u00e9es cart\u00e9siennes, qui utilisent diff\u00e9rents syst\u00e8mes de coordonn\u00e9es pour diriger les bras de la machine.", "token2charspan": [[0, 2], [3, 15], [16, 19], [20, 22], [23, 26], [27, 33], [34, 36], [37, 42], [43, 45], [46, 48], [49, 57], [57, 58], [59, 61], [62, 64], [65, 75], [76, 78], [79, 82], [83, 91], [92, 96], [97, 101], [102, 109], [110, 112], [113, 122], [123, 128], [129, 137], [137, 138], [139, 143], [144, 147], [148, 150], [151, 156], [157, 159], [160, 162], [163, 168], [169, 170], [171, 182], [183, 195], [195, 196], [197, 200], [201, 210], [211, 221], [222, 230], [231, 233], [234, 245], [246, 250], [251, 258], [259, 262], [263, 267], [268, 270], [271, 273], [274, 281], [281, 282]]}
{"doc_key": "ai-dev-294", "ner": [[1, 1, "country"], [8, 11, "organisation"], [15, 20, "organisation"], [24, 27, "organisation"], [31, 33, "organisation"], [37, 43, "organisation"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[8, 11, 1, 1, "physical", "", false, false], [15, 20, 1, 1, "physical", "", false, false], [24, 27, 1, 1, "physical", "", false, false], [31, 33, 1, 1, "physical", "", false, false], [37, 43, 1, 1, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Aux", "\u00c9tats-Unis", ",", "il", "est", "membre", "de", "la", "National", "Academy", "of", "Sciences", ",", "de", "l'", "American", "Academy", "of", "Arts", "and", "Sciences", ",", "de", "la", "Linguistic", "Society", "of", "America", ",", "de", "l'", "American", "Philosophical", "Association", "et", "de", "l'", "American", "Association", "for", "the", "Advancement", "of", "Science", "."], "sentence-detokenized": "Aux \u00c9tats-Unis, il est membre de la National Academy of Sciences, de l'American Academy of Arts and Sciences, de la Linguistic Society of America, de l'American Philosophical Association et de l'American Association for the Advancement of Science.", "token2charspan": [[0, 3], [4, 14], [14, 15], [16, 18], [19, 22], [23, 29], [30, 32], [33, 35], [36, 44], [45, 52], [53, 55], [56, 64], [64, 65], [66, 68], [69, 71], [71, 79], [80, 87], [88, 90], [91, 95], [96, 99], [100, 108], [108, 109], [110, 112], [113, 115], [116, 126], [127, 134], [135, 137], [138, 145], [145, 146], [147, 149], [150, 152], [152, 160], [161, 174], [175, 186], [187, 189], [190, 192], [193, 195], [195, 203], [204, 215], [216, 219], [220, 223], [224, 235], [236, 238], [239, 246], [246, 247]]}
{"doc_key": "ai-dev-295", "ner": [[11, 15, "algorithm"], [17, 17, "algorithm"], [26, 27, "algorithm"], [34, 35, "algorithm"], [42, 46, "task"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[11, 15, 26, 27, "named", "", false, false], [17, 17, 11, 15, "named", "", false, false], [26, 27, 34, 35, "compare", "", false, false], [26, 27, 42, 46, "related-to", "performs", false, false], [34, 35, 42, 46, "related-to", "performs", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Ils", "ont", "connu", "une", "grande", "notori\u00e9t\u00e9", "avec", "la", "popularit\u00e9", "de", "la", "machine", "\u00e0", "vecteur", "de", "support", "(", "SVM", ")", "dans", "les", "ann\u00e9es", "1990", ",", "lorsque", "la", "SVM", "s'", "est", "av\u00e9r\u00e9e", "comp\u00e9titive", "par", "rapport", "aux", "r\u00e9seaux", "neuronaux", "dans", "des", "t\u00e2ches", "telles", "que", "la", "reconnaissance", "de", "l'", "\u00e9criture", "manuscrite", "."], "sentence-detokenized": "Ils ont connu une grande notori\u00e9t\u00e9 avec la popularit\u00e9 de la machine \u00e0 vecteur de support (SVM) dans les ann\u00e9es 1990, lorsque la SVM s'est av\u00e9r\u00e9e comp\u00e9titive par rapport aux r\u00e9seaux neuronaux dans des t\u00e2ches telles que la reconnaissance de l'\u00e9criture manuscrite.", "token2charspan": [[0, 3], [4, 7], [8, 13], [14, 17], [18, 24], [25, 34], [35, 39], [40, 42], [43, 53], [54, 56], [57, 59], [60, 67], [68, 69], [70, 77], [78, 80], [81, 88], [89, 90], [90, 93], [93, 94], [95, 99], [100, 103], [104, 110], [111, 115], [115, 116], [117, 124], [125, 127], [128, 131], [132, 134], [134, 137], [138, 144], [145, 156], [157, 160], [161, 168], [169, 172], [173, 180], [181, 190], [191, 195], [196, 199], [200, 206], [207, 213], [214, 217], [218, 220], [221, 235], [236, 238], [239, 241], [241, 249], [250, 260], [260, 261]]}
{"doc_key": "ai-dev-296", "ner": [[0, 3, "misc"], [10, 10, "misc"], [16, 18, "algorithm"], [25, 27, "misc"], [35, 37, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[0, 3, 10, 10, "usage", "", false, false], [0, 3, 25, 27, "usage", "", false, false], [10, 10, 16, 18, "origin", "result_of_algorithm", false, false], [25, 27, 35, 37, "origin", "result_of_algorithm", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Une", "transform\u00e9e", "de", "blanchiment", "empirique", "est", "obtenue", "en", "estimant", "la", "covariance", "(", "par", "exemple", ",", "par", "maximum", "de", "vraisemblance", ")", "et", "en", "construisant", "ensuite", "une", "matrice", "de", "blanchiment", "estim\u00e9e", "correspondante", "(", "par", "exemple", ",", "par", "d\u00e9composition", "de", "Cholesky", ")", "."], "sentence-detokenized": "Une transform\u00e9e de blanchiment empirique est obtenue en estimant la covariance (par exemple, par maximum de vraisemblance) et en construisant ensuite une matrice de blanchiment estim\u00e9e correspondante (par exemple, par d\u00e9composition de Cholesky).", "token2charspan": [[0, 3], [4, 15], [16, 18], [19, 30], [31, 40], [41, 44], [45, 52], [53, 55], [56, 64], [65, 67], [68, 78], [79, 80], [80, 83], [84, 91], [91, 92], [93, 96], [97, 104], [105, 107], [108, 121], [121, 122], [123, 125], [126, 128], [129, 141], [142, 149], [150, 153], [154, 161], [162, 164], [165, 176], [177, 184], [185, 199], [200, 201], [201, 204], [205, 212], [212, 213], [214, 217], [218, 231], [232, 234], [235, 243], [243, 244], [244, 245]]}
{"doc_key": "ai-dev-297", "ner": [[0, 0, "organisation"], [9, 11, "product"], [21, 23, "product"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[9, 11, 0, 0, "artifact", "", false, false], [21, 23, 0, 0, "artifact", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["IAI", "est", "le", "plus", "grand", "fabricant", "mondial", "de", "robots", "\u00e0", "coordonn\u00e9es", "cart\u00e9siennes", "et", "est", "un", "leader", "reconnu", "dans", "le", "domaine", "des", "robots", "SCARA", "\u00e0", "faible", "co\u00fbt", "et", "\u00e0", "haute", "performance", "."], "sentence-detokenized": "IAI est le plus grand fabricant mondial de robots \u00e0 coordonn\u00e9es cart\u00e9siennes et est un leader reconnu dans le domaine des robots SCARA \u00e0 faible co\u00fbt et \u00e0 haute performance.", "token2charspan": [[0, 3], [4, 7], [8, 10], [11, 15], [16, 21], [22, 31], [32, 39], [40, 42], [43, 49], [50, 51], [52, 63], [64, 76], [77, 79], [80, 83], [84, 86], [87, 93], [94, 101], [102, 106], [107, 109], [110, 117], [118, 121], [122, 128], [129, 134], [135, 136], [137, 143], [144, 148], [149, 151], [152, 153], [154, 159], [160, 171], [171, 172]]}
{"doc_key": "ai-dev-298", "ner": [[15, 17, "field"], [20, 22, "field"], [25, 26, "field"], [29, 31, "field"], [34, 35, "field"], [38, 40, "field"], [43, 43, "field"], [46, 46, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "relations": [], "relations_mapping_to_source": [], "sentence": ["L'", "analyse", "formelle", "de", "concepts", "trouve", "des", "applications", "pratiques", "dans", "des", "domaines", "tels", "que", "l'", "exploration", "de", "donn\u00e9es", ",", "l'", "exploration", "de", "textes", ",", "l'", "apprentissage", "automatique", ",", "la", "gestion", "des", "connaissances", ",", "le", "web", "s\u00e9mantique", ",", "le", "d\u00e9veloppement", "de", "logiciels", ",", "la", "chimie", "et", "la", "biologie", "."], "sentence-detokenized": "L'analyse formelle de concepts trouve des applications pratiques dans des domaines tels que l'exploration de donn\u00e9es, l'exploration de textes, l'apprentissage automatique, la gestion des connaissances, le web s\u00e9mantique, le d\u00e9veloppement de logiciels, la chimie et la biologie.", "token2charspan": [[0, 2], [2, 9], [10, 18], [19, 21], [22, 30], [31, 37], [38, 41], [42, 54], [55, 64], [65, 69], [70, 73], [74, 82], [83, 87], [88, 91], [92, 94], [94, 105], [106, 108], [109, 116], [116, 117], [118, 120], [120, 131], [132, 134], [135, 141], [141, 142], [143, 145], [145, 158], [159, 170], [170, 171], [172, 174], [175, 182], [183, 186], [187, 200], [200, 201], [202, 204], [205, 208], [209, 219], [219, 220], [221, 223], [224, 237], [238, 240], [241, 250], [250, 251], [252, 254], [255, 261], [262, 264], [265, 267], [268, 276], [276, 277]]}
{"doc_key": "ai-dev-299", "ner": [[1, 1, "field"], [4, 8, "field"], [12, 15, "field"], [22, 23, "field"], [38, 39, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[4, 8, 22, 23, "part-of", "", false, false], [4, 8, 38, 39, "topic", "", false, false], [12, 15, 4, 8, "named", "", false, false], [22, 23, 1, 1, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["En", "informatique", ",", "la", "th\u00e9orie", "de", "l'", "apprentissage", "computationnel", "(", "ou", "simplement", "th\u00e9orie", "de", "l'", "apprentissage", ")", "est", "un", "sous-domaine", "de", "l'", "intelligence", "artificielle", "consacr\u00e9", "\u00e0", "l'", "\u00e9tude", "de", "la", "conception", "et", "de", "l'", "analyse", "des", "algorithmes", "d'", "apprentissage", "automatique", "."], "sentence-detokenized": "En informatique, la th\u00e9orie de l'apprentissage computationnel (ou simplement th\u00e9orie de l'apprentissage) est un sous-domaine de l'intelligence artificielle consacr\u00e9 \u00e0 l'\u00e9tude de la conception et de l'analyse des algorithmes d'apprentissage automatique.", "token2charspan": [[0, 2], [3, 15], [15, 16], [17, 19], [20, 27], [28, 30], [31, 33], [33, 46], [47, 61], [62, 63], [63, 65], [66, 76], [77, 84], [85, 87], [88, 90], [90, 103], [103, 104], [105, 108], [109, 111], [112, 124], [125, 127], [128, 130], [130, 142], [143, 155], [156, 164], [165, 166], [167, 169], [169, 174], [175, 177], [178, 180], [181, 191], [192, 194], [195, 197], [198, 200], [200, 207], [208, 211], [212, 223], [224, 226], [226, 239], [240, 251], [251, 252]]}
{"doc_key": "ai-dev-300", "ner": [[0, 2, "algorithm"], [4, 4, "algorithm"], [13, 14, "product"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[4, 4, 0, 2, "named", "", false, false], [13, 14, 0, 2, "usage", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Le", "filtrage", "collaboratif", "(", "FC", ")", "est", "une", "technique", "utilis\u00e9e", "par", "les", "syst\u00e8mes", "de", "recommandation", "."], "sentence-detokenized": "Le filtrage collaboratif (FC) est une technique utilis\u00e9e par les syst\u00e8mes de recommandation.", "token2charspan": [[0, 2], [3, 11], [12, 24], [25, 26], [26, 28], [28, 29], [30, 33], [34, 37], [38, 47], [48, 56], [57, 60], [61, 64], [65, 73], [74, 76], [77, 91], [91, 92]]}
{"doc_key": "ai-dev-301", "ner": [[0, 4, "metrics"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Le", "taux", "de", "FAUX", "positifs", "est", "la", "proportion", "de", "tous", "les", "n\u00e9gatifs", "qui", "donnent", "quand", "m\u00eame", "des", "r\u00e9sultats", "positifs", ",", "c'est-\u00e0-dire", "la", "probabilit\u00e9", "conditionnelle", "d'", "un", "r\u00e9sultat", "positif", "compte", "tenu", "d'", "un", "\u00e9v\u00e9nement", "qui", "n'", "\u00e9tait", "pas", "pr\u00e9sent", "."], "sentence-detokenized": "Le taux de FAUX positifs est la proportion de tous les n\u00e9gatifs qui donnent quand m\u00eame des r\u00e9sultats positifs, c'est-\u00e0-dire la probabilit\u00e9 conditionnelle d'un r\u00e9sultat positif compte tenu d'un \u00e9v\u00e9nement qui n'\u00e9tait pas pr\u00e9sent.", "token2charspan": [[0, 2], [3, 7], [8, 10], [11, 15], [16, 24], [25, 28], [29, 31], [32, 42], [43, 45], [46, 50], [51, 54], [55, 63], [64, 67], [68, 75], [76, 81], [82, 86], [87, 90], [91, 100], [101, 109], [109, 110], [111, 123], [124, 126], [127, 138], [139, 153], [154, 156], [156, 158], [159, 167], [168, 175], [176, 182], [183, 187], [188, 190], [190, 192], [193, 202], [203, 206], [207, 209], [209, 214], [215, 218], [219, 226], [226, 227]]}
{"doc_key": "ai-dev-302", "ner": [[1, 14, "misc"], [38, 38, "metrics"], [43, 43, "metrics"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[1, 14, 38, 38, "topic", "", false, false], [1, 14, 43, 43, "topic", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Dans", "VLDB", "'8", ":", "Proceedings", "of", "the", "34th", "International", "Conference", "on", "Very", "Large", "Data", "Bases", ",", "pages", "422", "--", "433", ".", "a", "montr\u00e9", "que", "les", "valeurs", "donn\u00e9es", "pour", "mathC", "/", "math", "et", "mathK", "/", "math", "impliquent", "g\u00e9n\u00e9ralement", "une", "pr\u00e9cision", "relativement", "faible", "des", "scores", "SimRank", "calcul\u00e9s", "par", "it\u00e9ration", "."], "sentence-detokenized": "Dans VLDB '8 : Proceedings of the 34th International Conference on Very Large Data Bases, pages 422--433. a montr\u00e9 que les valeurs donn\u00e9es pour mathC / math et mathK / math impliquent g\u00e9n\u00e9ralement une pr\u00e9cision relativement faible des scores SimRank calcul\u00e9s par it\u00e9ration.", "token2charspan": [[0, 4], [5, 9], [10, 12], [13, 14], [15, 26], [27, 29], [30, 33], [34, 38], [39, 52], [53, 63], [64, 66], [67, 71], [72, 77], [78, 82], [83, 88], [88, 89], [90, 95], [96, 99], [99, 101], [101, 104], [104, 105], [106, 107], [108, 114], [115, 118], [119, 122], [123, 130], [131, 138], [139, 143], [144, 149], [150, 151], [152, 156], [157, 159], [160, 165], [166, 167], [168, 172], [173, 183], [184, 196], [197, 200], [201, 210], [211, 223], [224, 230], [231, 234], [235, 241], [242, 249], [250, 258], [259, 262], [263, 272], [272, 273]]}
{"doc_key": "ai-dev-303", "ner": [[1, 3, "misc"], [4, 4, "misc"], [19, 19, "person"], [21, 23, "person"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[4, 4, 1, 3, "general-affiliation", "", false, false], [4, 4, 19, 19, "artifact", "", false, false], [4, 4, 21, 23, "artifact", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["Le", "drame", "de", "science-fiction", "Sense8", "a", "d\u00e9but\u00e9", "en", "juin", "2015", ".", "Il", "a", "\u00e9t\u00e9", "\u00e9crit", "et", "produit", "par", "les", "Wachowski", "et", "J.", "Michael", "Straczynski", "."], "sentence-detokenized": "Le drame de science-fiction Sense8 a d\u00e9but\u00e9 en juin 2015. Il a \u00e9t\u00e9 \u00e9crit et produit par les Wachowski et J. Michael Straczynski.", "token2charspan": [[0, 2], [3, 8], [9, 11], [12, 27], [28, 34], [35, 36], [37, 43], [44, 46], [47, 51], [52, 56], [56, 57], [58, 60], [61, 62], [63, 66], [67, 72], [73, 75], [76, 83], [84, 87], [88, 91], [92, 101], [102, 104], [105, 107], [108, 115], [116, 127], [127, 128]]}
{"doc_key": "ai-dev-304", "ner": [[2, 2, "misc"], [8, 10, "product"], [30, 35, "misc"], [46, 46, "country"], [50, 50, "country"], [54, 54, "country"], [57, 57, "country"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[2, 2, 8, 10, "topic", "", false, false], [46, 46, 30, 35, "type-of", "", false, false], [50, 50, 30, 35, "type-of", "", false, false], [54, 54, 30, 35, "type-of", "", false, false], [57, 57, 30, 35, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Bien", "qu'", "Eurotra", "n'", "ait", "jamais", "livr\u00e9", "un", "syst\u00e8me", "de", "TA", "op\u00e9rationnel", ",", "le", "projet", "a", "eu", "un", "impact", "consid\u00e9rable", "\u00e0", "long", "terme", "sur", "les", "industries", "langagi\u00e8res", "naissantes", "dans", "les", "\u00c9tats", "membres", "de", "l'", "Union", "europ\u00e9enne", ",", "en", "particulier", "dans", "les", "pays", "du", "sud", "de", "la", "Gr\u00e8ce", ",", "de", "l'", "Italie", ",", "de", "l'", "Espagne", "et", "du", "Portugal", "."], "sentence-detokenized": "Bien qu'Eurotra n'ait jamais livr\u00e9 un syst\u00e8me de TA op\u00e9rationnel, le projet a eu un impact consid\u00e9rable \u00e0 long terme sur les industries langagi\u00e8res naissantes dans les \u00c9tats membres de l'Union europ\u00e9enne, en particulier dans les pays du sud de la Gr\u00e8ce, de l'Italie, de l'Espagne et du Portugal.", "token2charspan": [[0, 4], [5, 8], [8, 15], [16, 18], [18, 21], [22, 28], [29, 34], [35, 37], [38, 45], [46, 48], [49, 51], [52, 64], [64, 65], [66, 68], [69, 75], [76, 77], [78, 80], [81, 83], [84, 90], [91, 103], [104, 105], [106, 110], [111, 116], [117, 120], [121, 124], [125, 135], [136, 147], [148, 158], [159, 163], [164, 167], [168, 173], [174, 181], [182, 184], [185, 187], [187, 192], [193, 203], [203, 204], [205, 207], [208, 219], [220, 224], [225, 228], [229, 233], [234, 236], [237, 240], [241, 243], [244, 246], [247, 252], [252, 253], [254, 256], [257, 259], [259, 265], [265, 266], [267, 269], [270, 272], [272, 279], [280, 282], [283, 285], [286, 294], [294, 295]]}
{"doc_key": "ai-dev-305", "ner": [[0, 1, "algorithm"], [9, 10, "task"], [17, 19, "task"], [21, 21, "task"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[9, 10, 0, 1, "usage", "", true, false], [17, 19, 9, 10, "named", "", false, false], [21, 21, 17, 19, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["L'", "autoencodeur", "a", "\u00e9t\u00e9", "appliqu\u00e9", "avec", "succ\u00e8s", "\u00e0", "la", "traduction", "automatique", "des", "langues", "humaines", ",", "g\u00e9n\u00e9ralement", "appel\u00e9e", "traduction", "automatique", "neuronale", "(", "NMT", ")", "."], "sentence-detokenized": "L'autoencodeur a \u00e9t\u00e9 appliqu\u00e9 avec succ\u00e8s \u00e0 la traduction automatique des langues humaines, g\u00e9n\u00e9ralement appel\u00e9e traduction automatique neuronale (NMT).", "token2charspan": [[0, 2], [2, 14], [15, 16], [17, 20], [21, 29], [30, 34], [35, 41], [42, 43], [44, 46], [47, 57], [58, 69], [70, 73], [74, 81], [82, 90], [90, 91], [92, 104], [105, 112], [113, 123], [124, 135], [136, 145], [146, 147], [147, 150], [150, 151], [151, 152]]}
{"doc_key": "ai-dev-306", "ner": [[15, 19, "metrics"], [22, 24, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Parmi", "les", "exemples", "populaires", "de", "fonctions", "d'", "aptitude", "bas\u00e9es", "sur", "les", "probabilit\u00e9s", ",", "citons", "l'", "estimation", "du", "maximum", "de", "vraisemblance", "et", "la", "perte", "de", "charni\u00e8re", "."], "sentence-detokenized": "Parmi les exemples populaires de fonctions d'aptitude bas\u00e9es sur les probabilit\u00e9s, citons l'estimation du maximum de vraisemblance et la perte de charni\u00e8re.", "token2charspan": [[0, 5], [6, 9], [10, 18], [19, 29], [30, 32], [33, 42], [43, 45], [45, 53], [54, 60], [61, 64], [65, 68], [69, 81], [81, 82], [83, 89], [90, 92], [92, 102], [103, 105], [106, 113], [114, 116], [117, 130], [131, 133], [134, 136], [137, 142], [143, 145], [146, 155], [155, 156]]}
{"doc_key": "ai-dev-307", "ner": [[0, 3, "field"], [14, 17, "task"], [23, 25, "field"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[14, 17, 0, 3, "part-of", "", false, false], [23, 25, 0, 3, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["L'", "exploration", "de", "donn\u00e9es", "est", "un", "domaine", "d'", "\u00e9tude", "connexe", ",", "ax\u00e9", "sur", "l'", "analyse", "exploratoire", "des", "donn\u00e9es", "par", "le", "biais", "d'", "un", "apprentissage", "non", "supervis\u00e9", "."], "sentence-detokenized": "L'exploration de donn\u00e9es est un domaine d'\u00e9tude connexe, ax\u00e9 sur l'analyse exploratoire des donn\u00e9es par le biais d'un apprentissage non supervis\u00e9.", "token2charspan": [[0, 2], [2, 13], [14, 16], [17, 24], [25, 28], [29, 31], [32, 39], [40, 42], [42, 47], [48, 55], [55, 56], [57, 60], [61, 64], [65, 67], [67, 74], [75, 87], [88, 91], [92, 99], [100, 103], [104, 106], [107, 112], [113, 115], [115, 117], [118, 131], [132, 135], [136, 145], [145, 146]]}
{"doc_key": "ai-dev-308", "ner": [[0, 2, "algorithm"], [20, 22, "product"]], "ner_mapping_to_source": [0, 1], "relations": [[0, 2, 20, 22, "related-to", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Le", "filtrage", "collaboratif", "englobe", "des", "techniques", "permettant", "de", "faire", "correspondre", "des", "personnes", "ayant", "des", "int\u00e9r\u00eats", "similaires", "et", "de", "cr\u00e9er", "un", "syst\u00e8me", "de", "recommandation", "sur", "cette", "base", "."], "sentence-detokenized": "Le filtrage collaboratif englobe des techniques permettant de faire correspondre des personnes ayant des int\u00e9r\u00eats similaires et de cr\u00e9er un syst\u00e8me de recommandation sur cette base.", "token2charspan": [[0, 2], [3, 11], [12, 24], [25, 32], [33, 36], [37, 47], [48, 58], [59, 61], [62, 67], [68, 80], [81, 84], [85, 94], [95, 100], [101, 104], [105, 113], [114, 124], [125, 127], [128, 130], [131, 136], [137, 139], [140, 147], [148, 150], [151, 165], [166, 169], [170, 175], [176, 180], [180, 181]]}
{"doc_key": "ai-dev-309", "ner": [[4, 11, "algorithm"], [19, 19, "programlang"], [21, 24, "product"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[21, 24, 4, 11, "type-of", "", false, false], [21, 24, 19, 19, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Un", "certain", "nombre", "d'", "algorithmes", "de", "similarit\u00e9", "de", "mots", "bas\u00e9s", "sur", "WordNet", "sont", "mis", "en", "\u0153uvre", "dans", "un", "paquetage", "Perl", "appel\u00e9", "WordNet", ":", ":", "Similarity", "."], "sentence-detokenized": "Un certain nombre d'algorithmes de similarit\u00e9 de mots bas\u00e9s sur WordNet sont mis en \u0153uvre dans un paquetage Perl appel\u00e9 WordNet: : Similarity.", "token2charspan": [[0, 2], [3, 10], [11, 17], [18, 20], [20, 31], [32, 34], [35, 45], [46, 48], [49, 53], [54, 59], [60, 63], [64, 71], [72, 76], [77, 80], [81, 83], [84, 89], [90, 94], [95, 97], [98, 107], [108, 112], [113, 119], [120, 127], [127, 128], [129, 130], [131, 141], [141, 142]]}
{"doc_key": "ai-dev-310", "ner": [[7, 7, "conference"], [9, 9, "conference"], [13, 14, "researcher"], [16, 17, "researcher"], [19, 22, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[9, 9, 7, 7, "named", "", false, false], [13, 14, 7, 7, "temporal", "", false, false], [16, 17, 7, 7, "temporal", "", false, false], [19, 22, 7, 7, "temporal", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["Un", "autre", "article", ",", "pr\u00e9sent\u00e9", "\u00e0", "la", "CVPR", "(", "CVPR", ")", "2000", "par", "Erik", "Miller", ",", "Nicholas", "Matsakis", "et", "Paul", "Viola", ",", "sera", "\u00e9galement", "discut\u00e9", "."], "sentence-detokenized": "Un autre article, pr\u00e9sent\u00e9 \u00e0 la CVPR (CVPR) 2000 par Erik Miller, Nicholas Matsakis et Paul Viola, sera \u00e9galement discut\u00e9.", "token2charspan": [[0, 2], [3, 8], [9, 16], [16, 17], [18, 26], [27, 28], [29, 31], [32, 36], [37, 38], [38, 42], [42, 43], [44, 48], [49, 52], [53, 57], [58, 64], [64, 65], [66, 74], [75, 83], [84, 86], [87, 91], [92, 97], [97, 98], [99, 103], [104, 113], [114, 121], [121, 122]]}
{"doc_key": "ai-dev-311", "ner": [[0, 1, "algorithm"], [10, 12, "misc"], [21, 23, "metrics"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[0, 1, 21, 23, "compare", "", false, false], [21, 23, 10, 12, "type-of", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Le", "CQ", "n'", "a", "pas", "\u00e9t\u00e9", "\u00e9valu\u00e9", "par", "rapport", "aux", "algorithmes", "de", "regroupement", "modernes", "traditionnels", ",", "\u00e0", "l'", "exception", "de", "l'", "indice", "de", "Jaccard", "."], "sentence-detokenized": "Le CQ n'a pas \u00e9t\u00e9 \u00e9valu\u00e9 par rapport aux algorithmes de regroupement modernes traditionnels, \u00e0 l'exception de l'indice de Jaccard.", "token2charspan": [[0, 2], [3, 5], [6, 8], [8, 9], [10, 13], [14, 17], [18, 24], [25, 28], [29, 36], [37, 40], [41, 52], [53, 55], [56, 68], [69, 77], [78, 91], [91, 92], [93, 94], [95, 97], [97, 106], [107, 109], [110, 112], [112, 118], [119, 121], [122, 129], [129, 130]]}
{"doc_key": "ai-dev-312", "ner": [[2, 7, "misc"], [10, 12, "misc"], [17, 18, "location"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[10, 12, 2, 7, "physical", "", false, false], [10, 12, 17, 18, "temporal", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Pendant", "le", "championnat", "du", "monde", "de", "robotique", "VEX", ",", "un", "d\u00e9fil\u00e9", "des", "nations", "est", "organis\u00e9", "dans", "le", "Freedom", "Hall", ",", "auquel", "participent", "des", "centaines", "d'", "\u00e9tudiants", "de", "plus", "de", "30", "pays", "."], "sentence-detokenized": "Pendant le championnat du monde de robotique VEX, un d\u00e9fil\u00e9 des nations est organis\u00e9 dans le Freedom Hall, auquel participent des centaines d'\u00e9tudiants de plus de 30 pays.", "token2charspan": [[0, 7], [8, 10], [11, 22], [23, 25], [26, 31], [32, 34], [35, 44], [45, 48], [48, 49], [50, 52], [53, 59], [60, 63], [64, 71], [72, 75], [76, 84], [85, 89], [90, 92], [93, 100], [101, 105], [105, 106], [107, 113], [114, 125], [126, 129], [130, 139], [140, 142], [142, 151], [152, 154], [155, 159], [160, 162], [163, 165], [166, 170], [170, 171]]}
{"doc_key": "ai-dev-313", "ner": [[8, 14, "metrics"], [16, 16, "metrics"], [20, 24, "metrics"], [26, 26, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [[16, 16, 8, 14, "named", "", false, false], [26, 26, 20, 24, "named", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["D'", "autres", "mesures", "de", "la", "pr\u00e9cision", "comprennent", "le", "taux", "d'", "erreur", "sur", "un", "seul", "mot", "(", "SWER", ")", "et", "le", "taux", "de", "r\u00e9ussite", "des", "commandes", "(", "CSR", ")", "."], "sentence-detokenized": "D'autres mesures de la pr\u00e9cision comprennent le taux d'erreur sur un seul mot (SWER) et le taux de r\u00e9ussite des commandes (CSR).", "token2charspan": [[0, 2], [2, 8], [9, 16], [17, 19], [20, 22], [23, 32], [33, 44], [45, 47], [48, 52], [53, 55], [55, 61], [62, 65], [66, 68], [69, 73], [74, 77], [78, 79], [79, 83], [83, 84], [85, 87], [88, 90], [91, 95], [96, 98], [99, 107], [108, 111], [112, 121], [122, 123], [123, 126], [126, 127], [127, 128]]}
{"doc_key": "ai-dev-314", "ner": [[9, 10, "conference"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Ils", "ont", "pr\u00e9sent\u00e9", "leur", "m\u00e9thode", "et", "leurs", "r\u00e9sultats", "au", "SIGGRAPH", "2000", "."], "sentence-detokenized": "Ils ont pr\u00e9sent\u00e9 leur m\u00e9thode et leurs r\u00e9sultats au SIGGRAPH 2000.", "token2charspan": [[0, 3], [4, 7], [8, 16], [17, 21], [22, 29], [30, 32], [33, 38], [39, 48], [49, 51], [52, 60], [61, 65], [65, 66]]}
{"doc_key": "ai-dev-315", "ner": [[0, 2, "conference"], [7, 7, "misc"], [9, 13, "misc"], [17, 21, "conference"], [28, 30, "researcher"], [40, 41, "researcher"], [45, 47, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[0, 2, 7, 7, "origin", "", false, false], [7, 7, 17, 21, "physical", "", false, false], [7, 7, 17, 21, "temporal", "", false, false], [7, 7, 28, 30, "origin", "", false, false], [7, 7, 40, 41, "origin", "", false, false], [9, 13, 7, 7, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["La", "conf\u00e9rence", "KDD", "est", "issue", "des", "ateliers", "KDD", "(", "Knowledge", "Discovery", "and", "Data", "Mining", ")", "organis\u00e9s", "lors", "des", "conf\u00e9rences", "de", "l'", "AAAI", ",", "qui", "ont", "\u00e9t\u00e9", "lanc\u00e9s", "par", "Gregory", "I.", "Piatetsky-Shapiro", "en", "1989", ",", "1991", "et", "1993", ",", "et", "par", "Usama", "Fayyad", "en", "1994", ".", "Machinery", "|", "ACM."], "sentence-detokenized": "La conf\u00e9rence KDD est issue des ateliers KDD (Knowledge Discovery and Data Mining) organis\u00e9s lors des conf\u00e9rences de l'AAAI, qui ont \u00e9t\u00e9 lanc\u00e9s par Gregory I. Piatetsky-Shapiro en 1989, 1991 et 1993, et par Usama Fayyad en 1994. Machinery | ACM.", "token2charspan": [[0, 2], [3, 13], [14, 17], [18, 21], [22, 27], [28, 31], [32, 40], [41, 44], [45, 46], [46, 55], [56, 65], [66, 69], [70, 74], [75, 81], [81, 82], [83, 92], [93, 97], [98, 101], [102, 113], [114, 116], [117, 119], [119, 123], [123, 124], [125, 128], [129, 132], [133, 136], [137, 143], [144, 147], [148, 155], [156, 158], [159, 176], [177, 179], [180, 184], [184, 185], [186, 190], [191, 193], [194, 198], [198, 199], [200, 202], [203, 206], [207, 212], [213, 219], [220, 222], [223, 227], [227, 228], [229, 238], [239, 240], [241, 245]]}
{"doc_key": "ai-dev-316", "ner": [[7, 10, "conference"], [12, 12, "conference"], [17, 22, "organisation"], [24, 24, "organisation"], [29, 33, "conference"], [35, 35, "conference"], [40, 46, "conference"], [48, 48, "conference"], [53, 58, "conference"], [60, 60, "conference"], [65, 70, "conference"], [72, 72, "conference"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], "relations": [[12, 12, 7, 10, "named", "", false, false], [24, 24, 17, 22, "named", "", false, false], [35, 35, 29, 33, "named", "", false, false], [48, 48, 40, 46, "named", "", false, false], [60, 60, 53, 58, "named", "", false, false], [72, 72, 65, 70, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5], "sentence": ["Il", "a", "\u00e9t\u00e9", "\u00e9lu", "membre", "de", "l'", "Association", "for", "Computing", "Machinery", "(", "ACM", ")", ",", "de", "l'", "Institute", "of", "Electrical", "and", "Electronics", "Engineers", "(", "IEEE", ")", ",", "de", "l'", "International", "Association", "for", "Pattern", "Recognition", "(", "IAPR", ")", ",", "de", "l'", "Association", "for", "the", "Advancement", "of", "Artificial", "Intelligence", "(", "AAAI", ")", ",", "de", "l'", "American", "Association", "for", "Advancement", "of", "Science", "(", "AAAS", ")", "et", "de", "la", "Society", "for", "Optics", "and", "Photonics", "Technology", "(", "SPIE", ")", "."], "sentence-detokenized": "Il a \u00e9t\u00e9 \u00e9lu membre de l'Association for Computing Machinery (ACM), de l'Institute of Electrical and Electronics Engineers (IEEE), de l'International Association for Pattern Recognition (IAPR), de l'Association for the Advancement of Artificial Intelligence (AAAI), de l'American Association for Advancement of Science (AAAS) et de la Society for Optics and Photonics Technology (SPIE).", "token2charspan": [[0, 2], [3, 4], [5, 8], [9, 12], [13, 19], [20, 22], [23, 25], [25, 36], [37, 40], [41, 50], [51, 60], [61, 62], [62, 65], [65, 66], [66, 67], [68, 70], [71, 73], [73, 82], [83, 85], [86, 96], [97, 100], [101, 112], [113, 122], [123, 124], [124, 128], [128, 129], [129, 130], [131, 133], [134, 136], [136, 149], [150, 161], [162, 165], [166, 173], [174, 185], [186, 187], [187, 191], [191, 192], [192, 193], [194, 196], [197, 199], [199, 210], [211, 214], [215, 218], [219, 230], [231, 233], [234, 244], [245, 257], [258, 259], [259, 263], [263, 264], [264, 265], [266, 268], [269, 271], [271, 279], [280, 291], [292, 295], [296, 307], [308, 310], [311, 318], [319, 320], [320, 324], [324, 325], [326, 328], [329, 331], [332, 334], [335, 342], [343, 346], [347, 353], [354, 357], [358, 367], [368, 378], [379, 380], [380, 384], [384, 385], [385, 386]]}
{"doc_key": "ai-dev-317", "ner": [[0, 2, "field"], [5, 7, "field"], [24, 25, "field"], [47, 49, "field"], [73, 80, "task"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[0, 2, 24, 25, "named", "", false, false], [5, 7, 47, 49, "named", "", false, false], [47, 49, 73, 80, "related-to", "", true, false]], "relations_mapping_to_source": [0, 1, 2], "sentence": ["L'", "apprentissage", "automatique", "et", "l'", "exploration", "de", "donn\u00e9es", "utilisent", "souvent", "les", "m\u00eames", "m\u00e9thodes", "et", "se", "chevauchent", "de", "mani\u00e8re", "significative", ",", "mais", "alors", "que", "l'", "apprentissage", "automatique", "se", "concentre", "sur", "la", "pr\u00e9diction", ",", "sur", "la", "base", "de", "propri\u00e9t\u00e9s", "connues", "apprises", "\u00e0", "partir", "des", "donn\u00e9es", "d'", "apprentissage", ",", "l'", "exploration", "de", "donn\u00e9es", "se", "concentre", "sur", "la", "d\u00e9couverte", "de", "propri\u00e9t\u00e9s", "(", "pr\u00e9c\u00e9demment", ")", "inconnues", "dans", "les", "donn\u00e9es", "(", "c'", "est", "l'", "\u00e9tape", "d'", "analyse", "de", "la", "d\u00e9couverte", "de", "connaissances", "dans", "les", "bases", "de", "donn\u00e9es", ")", "."], "sentence-detokenized": "L'apprentissage automatique et l'exploration de donn\u00e9es utilisent souvent les m\u00eames m\u00e9thodes et se chevauchent de mani\u00e8re significative, mais alors que l'apprentissage automatique se concentre sur la pr\u00e9diction, sur la base de propri\u00e9t\u00e9s connues apprises \u00e0 partir des donn\u00e9es d'apprentissage, l'exploration de donn\u00e9es se concentre sur la d\u00e9couverte de propri\u00e9t\u00e9s (pr\u00e9c\u00e9demment) inconnues dans les donn\u00e9es (c'est l'\u00e9tape d'analyse de la d\u00e9couverte de connaissances dans les bases de donn\u00e9es).", "token2charspan": [[0, 2], [2, 15], [16, 27], [28, 30], [31, 33], [33, 44], [45, 47], [48, 55], [56, 65], [66, 73], [74, 77], [78, 83], [84, 92], [93, 95], [96, 98], [99, 110], [111, 113], [114, 121], [122, 135], [135, 136], [137, 141], [142, 147], [148, 151], [152, 154], [154, 167], [168, 179], [180, 182], [183, 192], [193, 196], [197, 199], [200, 210], [210, 211], [212, 215], [216, 218], [219, 223], [224, 226], [227, 237], [238, 245], [246, 254], [255, 256], [257, 263], [264, 267], [268, 275], [276, 278], [278, 291], [291, 292], [293, 295], [295, 306], [307, 309], [310, 317], [318, 320], [321, 330], [331, 334], [335, 337], [338, 348], [349, 351], [352, 362], [363, 364], [364, 376], [376, 377], [378, 387], [388, 392], [393, 396], [397, 404], [405, 406], [406, 408], [408, 411], [412, 414], [414, 419], [420, 422], [422, 429], [430, 432], [433, 435], [436, 446], [447, 449], [450, 463], [464, 468], [469, 472], [473, 478], [479, 481], [482, 489], [489, 490], [490, 491]]}
{"doc_key": "ai-dev-318", "ner": [[0, 0, "product"], [4, 4, "programlang"], [14, 14, "misc"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[0, 0, 4, 4, "general-affiliation", "", false, false], [0, 0, 4, 4, "related-to", "runs_on", true, false]], "relations_mapping_to_source": [0, 1], "sentence": ["Indy", "est", "\u00e9crit", "en", "Java", "et", "fonctionne", "donc", "sur", "la", "plupart", "des", "syst\u00e8mes", "d'", "exploitation", "modernes", "."], "sentence-detokenized": "Indy est \u00e9crit en Java et fonctionne donc sur la plupart des syst\u00e8mes d'exploitation modernes.", "token2charspan": [[0, 4], [5, 8], [9, 14], [15, 17], [18, 22], [23, 25], [26, 36], [37, 41], [42, 45], [46, 48], [49, 56], [57, 60], [61, 69], [70, 72], [72, 84], [85, 93], [93, 94]]}
{"doc_key": "ai-dev-319", "ner": [[0, 1, "algorithm"], [7, 10, "algorithm"], [12, 12, "algorithm"], [18, 22, "algorithm"], [24, 24, "algorithm"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[0, 1, 7, 10, "type-of", "", true, false], [12, 12, 7, 10, "named", "", false, false], [18, 22, 7, 10, "type-of", "", true, false], [24, 24, 18, 22, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3], "sentence": ["La", "NMF", "est", "une", "instance", "de", "la", "programmation", "quadratique", "non", "n\u00e9gative", "(", "NQP", ")", ",", "tout", "comme", "la", "machine", "\u00e0", "vecteurs", "de", "support", "(", "SVM", ")", "."], "sentence-detokenized": "La NMF est une instance de la programmation quadratique non n\u00e9gative (NQP), tout comme la machine \u00e0 vecteurs de support (SVM).", "token2charspan": [[0, 2], [3, 6], [7, 10], [11, 14], [15, 23], [24, 26], [27, 29], [30, 43], [44, 55], [56, 59], [60, 68], [69, 70], [70, 73], [73, 74], [74, 75], [76, 80], [81, 86], [87, 89], [90, 97], [98, 99], [100, 108], [109, 111], [112, 119], [120, 121], [121, 124], [124, 125], [125, 126]]}
{"doc_key": "ai-dev-320", "ner": [[7, 8, "misc"], [15, 20, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[7, 8, 15, 20, "usage", "", true, false]], "relations_mapping_to_source": [0], "sentence": ["La", "m\u00e9thode", "repose", "sur", "l'", "estimation", "des", "probabilit\u00e9s", "conditionnelles", "\u00e0", "l'", "aide", "de", "la", "m\u00e9thode", "non", "param\u00e9trique", "du", "maximum", "de", "vraisemblance", "qui", "conduit", "\u00e0"], "sentence-detokenized": "La m\u00e9thode repose sur l'estimation des probabilit\u00e9s conditionnelles \u00e0 l'aide de la m\u00e9thode non param\u00e9trique du maximum de vraisemblance qui conduit \u00e0", "token2charspan": [[0, 2], [3, 10], [11, 17], [18, 21], [22, 24], [24, 34], [35, 38], [39, 51], [52, 67], [68, 69], [70, 72], [72, 76], [77, 79], [80, 82], [83, 90], [91, 94], [95, 107], [108, 110], [111, 118], [119, 121], [122, 135], [136, 139], [140, 147], [148, 149]]}
{"doc_key": "ai-dev-321", "ner": [[10, 10, "algorithm"], [13, 16, "algorithm"], [19, 21, "metrics"], [24, 24, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [], "relations_mapping_to_source": [], "sentence": ["Les", "concepts", "de", "base", "de", "l'", "estimation", "spectrale", "comprennent", "l'", "autocorr\u00e9lation", ",", "la", "transform\u00e9e", "de", "Fourier", "multi-D", ",", "l'", "erreur", "quadratique", "moyenne", "et", "l'", "entropie", "."], "sentence-detokenized": "Les concepts de base de l'estimation spectrale comprennent l'autocorr\u00e9lation, la transform\u00e9e de Fourier multi-D, l'erreur quadratique moyenne et l'entropie.", "token2charspan": [[0, 3], [4, 12], [13, 15], [16, 20], [21, 23], [24, 26], [26, 36], [37, 46], [47, 58], [59, 61], [61, 76], [76, 77], [78, 80], [81, 92], [93, 95], [96, 103], [104, 111], [111, 112], [113, 115], [115, 121], [122, 133], [134, 141], [142, 144], [145, 147], [147, 155], [155, 156]]}
{"doc_key": "ai-dev-322", "ner": [[5, 7, "algorithm"], [13, 13, "field"], [16, 16, "algorithm"], [19, 23, "algorithm"], [26, 27, "task"], [30, 30, "field"], [33, 33, "field"], [36, 38, "task"], [41, 45, "task"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "relations": [[5, 7, 13, 13, "part-of", "", false, false], [5, 7, 16, 16, "part-of", "", false, false], [5, 7, 19, 23, "part-of", "", false, false], [5, 7, 26, 27, "part-of", "", false, false], [5, 7, 30, 30, "part-of", "", false, false], [5, 7, 33, 33, "part-of", "", false, false], [5, 7, 36, 38, "part-of", "", false, false], [5, 7, 41, 45, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7], "sentence": ["Les", "domaines", "d'", "application", "des", "m\u00e9thodes", "\u00e0", "noyau", "sont", "vari\u00e9s", "et", "comprennent", "la", "g\u00e9ostatistique", ",", "le", "krigeage", ",", "la", "pond\u00e9ration", "de", "la", "distance", "inverse", ",", "la", "reconstruction", "3D", ",", "la", "bioinformatique", ",", "la", "chimio-informatique", ",", "l'", "extraction", "d'", "informations", "et", "la", "reconnaissance", "de", "l'", "\u00e9criture", "manuscrite", "."], "sentence-detokenized": "Les domaines d'application des m\u00e9thodes \u00e0 noyau sont vari\u00e9s et comprennent la g\u00e9ostatistique, le krigeage, la pond\u00e9ration de la distance inverse, la reconstruction 3D, la bioinformatique, la chimio-informatique, l'extraction d'informations et la reconnaissance de l'\u00e9criture manuscrite.", "token2charspan": [[0, 3], [4, 12], [13, 15], [15, 26], [27, 30], [31, 39], [40, 41], [42, 47], [48, 52], [53, 59], [60, 62], [63, 74], [75, 77], [78, 92], [92, 93], [94, 96], [97, 105], [105, 106], [107, 109], [110, 121], [122, 124], [125, 127], [128, 136], [137, 144], [144, 145], [146, 148], [149, 163], [164, 166], [166, 167], [168, 170], [171, 186], [186, 187], [188, 190], [191, 210], [210, 211], [212, 214], [214, 224], [225, 227], [227, 239], [240, 242], [243, 245], [246, 260], [261, 263], [264, 266], [266, 274], [275, 285], [285, 286]]}
{"doc_key": "ai-dev-323", "ner": [[23, 23, "organisation"], [16, 20, "product"], [14, 14, "product"], [28, 35, "organisation"], [29, 34, "product"], [26, 26, "product"], [37, 38, "product"], [41, 42, "product"], [45, 49, "product"], [52, 55, "product"], [58, 60, "product"], [65, 65, "product"], [69, 73, "product"], [78, 80, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13], "relations": [[16, 20, 23, 23, "artifact", "", false, false], [16, 20, 37, 38, "compare", "", false, false], [16, 20, 41, 42, "compare", "", false, false], [16, 20, 45, 49, "compare", "", false, false], [16, 20, 52, 55, "compare", "", false, false], [16, 20, 58, 60, "compare", "", false, false], [16, 20, 65, 65, "compare", "", false, false], [16, 20, 69, 73, "compare", "", false, false], [16, 20, 78, 80, "compare", "", false, false], [14, 14, 16, 20, "named", "", false, false], [29, 34, 28, 35, "artifact", "", false, false], [29, 34, 37, 38, "compare", "", false, false], [29, 34, 41, 42, "compare", "", false, false], [29, 34, 45, 49, "compare", "", false, false], [29, 34, 52, 55, "compare", "", false, false], [29, 34, 58, 60, "compare", "", false, false], [29, 34, 65, 65, "compare", "", false, false], [29, 34, 69, 73, "compare", "", false, false], [29, 34, 78, 80, "compare", "", false, false], [26, 26, 29, 34, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19], "sentence": ["Les", "robots", "peuvent", "\u00eatre", "autonomes", "ou", "semi-autonomes", "et", "vont", "des", "humano\u00efdes", "tels", "que", "l'", "ASIMO", "(", "Advanced", "Step", "in", "Innovative", "Mobility", ")", "de", "Honda", "et", "le", "TOPIO", "(", "TOSY", "Ping", "Pong", "Playing", "Robot", ")", "de", "TOSY", "aux", "robots", "industriels", ",", "aux", "robots", "m\u00e9dicaux", ",", "aux", "robots", "d'", "assistance", "aux", "patients", ",", "aux", "robots", "de", "th\u00e9rapie", "canine", ",", "aux", "robots", "en", "essaim", "programm\u00e9s", "collectivement", ",", "aux", "drones", "tels", "que", "le", "MQ-1", "Predator", "de", "General", "Atomics", ",", "et", "m\u00eame", "aux", "nano", "robots", "microscopiques", "."], "sentence-detokenized": "Les robots peuvent \u00eatre autonomes ou semi-autonomes et vont des humano\u00efdes tels que l'ASIMO (Advanced Step in Innovative Mobility) de Honda et le TOPIO (TOSY Ping Pong Playing Robot) de TOSY aux robots industriels, aux robots m\u00e9dicaux, aux robots d'assistance aux patients, aux robots de th\u00e9rapie canine, aux robots en essaim programm\u00e9s collectivement, aux drones tels que le MQ-1 Predator de General Atomics, et m\u00eame aux nano robots microscopiques.", "token2charspan": [[0, 3], [4, 10], [11, 18], [19, 23], [24, 33], [34, 36], [37, 51], [52, 54], [55, 59], [60, 63], [64, 74], [75, 79], [80, 83], [84, 86], [86, 91], [92, 93], [93, 101], [102, 106], [107, 109], [110, 120], [121, 129], [129, 130], [131, 133], [134, 139], [140, 142], [143, 145], [146, 151], [152, 153], [153, 157], [158, 162], [163, 167], [168, 175], [176, 181], [181, 182], [183, 185], [186, 190], [191, 194], [195, 201], [202, 213], [213, 214], [215, 218], [219, 225], [226, 234], [234, 235], [236, 239], [240, 246], [247, 249], [249, 259], [260, 263], [264, 272], [272, 273], [274, 277], [278, 284], [285, 287], [288, 296], [297, 303], [303, 304], [305, 308], [309, 315], [316, 318], [319, 325], [326, 336], [337, 351], [351, 352], [353, 356], [357, 363], [364, 368], [369, 372], [373, 375], [376, 380], [381, 389], [390, 392], [393, 400], [401, 408], [408, 409], [410, 412], [413, 417], [418, 421], [422, 426], [427, 433], [434, 448], [448, 449]]}
{"doc_key": "ai-dev-324", "ner": [[0, 0, "product"], [2, 3, "product"], [10, 17, "university"], [19, 20, "researcher"], [22, 23, "researcher"], [25, 26, "researcher"], [28, 29, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[0, 0, 19, 20, "artifact", "", false, false], [0, 0, 22, 23, "artifact", "", false, false], [0, 0, 25, 26, "artifact", "", false, false], [0, 0, 28, 29, "artifact", "", false, false], [2, 3, 19, 20, "artifact", "", false, false], [2, 3, 22, 23, "artifact", "", false, false], [2, 3, 25, 26, "artifact", "", false, false], [2, 3, 28, 29, "artifact", "", false, false], [19, 20, 10, 17, "physical", "", false, false], [22, 23, 10, 17, "physical", "", false, false], [25, 26, 10, 17, "physical", "", false, false], [28, 29, 10, 17, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], "sentence": ["Freddy", "et", "Freddy", "II", "\u00e9taient", "des", "robots", "construits", "\u00e0", "l'", "\u00e9cole", "d'", "informatique", "de", "l'", "universit\u00e9", "d'", "\u00c9dimbourg", "par", "Pat", "Ambler", ",", "Robin", "Popplestone", ",", "Austin", "Tate", "et", "Donald", "Mitchie", ",", "et", "\u00e9taient", "capables", "d'", "assembler", "des", "blocs", "de", "bois", "en", "plusieurs", "heures", "."], "sentence-detokenized": "Freddy et Freddy II \u00e9taient des robots construits \u00e0 l'\u00e9cole d'informatique de l'universit\u00e9 d'\u00c9dimbourg par Pat Ambler, Robin Popplestone, Austin Tate et Donald Mitchie, et \u00e9taient capables d'assembler des blocs de bois en plusieurs heures.", "token2charspan": [[0, 6], [7, 9], [10, 16], [17, 19], [20, 27], [28, 31], [32, 38], [39, 49], [50, 51], [52, 54], [54, 59], [60, 62], [62, 74], [75, 77], [78, 80], [80, 90], [91, 93], [93, 102], [103, 106], [107, 110], [111, 117], [117, 118], [119, 124], [125, 136], [136, 137], [138, 144], [145, 149], [150, 152], [153, 159], [160, 167], [167, 168], [169, 171], [172, 179], [180, 188], [189, 191], [191, 200], [201, 204], [205, 210], [211, 213], [214, 218], [219, 221], [222, 231], [232, 238], [238, 239]]}
{"doc_key": "ai-dev-325", "ner": [[6, 6, "location"], [9, 9, "country"], [17, 17, "country"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[6, 6, 9, 9, "physical", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Il", "a", "pass\u00e9", "son", "enfance", "\u00e0", "Paris", ",", "en", "France", ",", "o\u00f9", "ses", "parents", "avaient", "\u00e9migr\u00e9", "de", "Lituanie", "au", "d\u00e9but", "des", "ann\u00e9es", "1920", "."], "sentence-detokenized": "Il a pass\u00e9 son enfance \u00e0 Paris, en France, o\u00f9 ses parents avaient \u00e9migr\u00e9 de Lituanie au d\u00e9but des ann\u00e9es 1920.", "token2charspan": [[0, 2], [3, 4], [5, 10], [11, 14], [15, 22], [23, 24], [25, 30], [30, 31], [32, 34], [35, 41], [41, 42], [43, 45], [46, 49], [50, 57], [58, 65], [66, 72], [73, 75], [76, 84], [85, 87], [88, 93], [94, 97], [98, 104], [105, 109], [109, 110]]}
{"doc_key": "ai-dev-326", "ner": [[2, 3, "researcher"], [8, 12, "misc"], [15, 17, "organisation"], [20, 22, "university"], [33, 36, "university"], [48, 50, "university"], [53, 57, "university"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[2, 3, 8, 12, "role", "", false, false], [2, 3, 20, 22, "physical", "", false, false], [2, 3, 33, 36, "role", "", false, false], [2, 3, 48, 50, "role", "", false, false], [2, 3, 53, 57, "role", "", false, false], [8, 12, 15, 17, "part-of", "", false, false], [15, 17, 20, 22, "part-of", "", false, false], [48, 50, 33, 36, "part-of", "", false, false], [53, 57, 33, 36, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "sentence": ["Auparavant", ",", "M.", "Paulos", "\u00e9tait", "titulaire", "de", "la", "chaire", "de", "professeur", "associ\u00e9", "Cooper-Siegel", "\u00e0", "l'", "\u00e9cole", "d'", "informatique", "de", "l'", "universit\u00e9", "Carnegie", "Mellon", ",", "o\u00f9", "il", "\u00e9tait", "membre", "du", "corps", "professoral", "de", "l'", "Institut", "d'", "interaction", "homme-machine", "et", "o\u00f9", "il", "\u00e9tait", "\u00e9galement", "membre", "du", "corps", "professoral", "de", "l'", "Institut", "de", "robotique", "et", "du", "Centre", "des", "technologies", "du", "divertissement", "."], "sentence-detokenized": "Auparavant, M. Paulos \u00e9tait titulaire de la chaire de professeur associ\u00e9 Cooper-Siegel \u00e0 l'\u00e9cole d'informatique de l'universit\u00e9 Carnegie Mellon, o\u00f9 il \u00e9tait membre du corps professoral de l'Institut d'interaction homme-machine et o\u00f9 il \u00e9tait \u00e9galement membre du corps professoral de l'Institut de robotique et du Centre des technologies du divertissement.", "token2charspan": [[0, 10], [10, 11], [12, 14], [15, 21], [22, 27], [28, 37], [38, 40], [41, 43], [44, 50], [51, 53], [54, 64], [65, 72], [73, 86], [87, 88], [89, 91], [91, 96], [97, 99], [99, 111], [112, 114], [115, 117], [117, 127], [128, 136], [137, 143], [143, 144], [145, 147], [148, 150], [151, 156], [157, 163], [164, 166], [167, 172], [173, 184], [185, 187], [188, 190], [190, 198], [199, 201], [201, 212], [213, 226], [227, 229], [230, 232], [233, 235], [236, 241], [242, 251], [252, 258], [259, 261], [262, 267], [268, 279], [280, 282], [283, 285], [285, 293], [294, 296], [297, 306], [307, 309], [310, 312], [313, 319], [320, 323], [324, 336], [337, 339], [340, 354], [354, 355]]}
{"doc_key": "ai-dev-327", "ner": [[3, 4, "researcher"], [7, 9, "university"], [13, 15, "product"], [18, 22, "product"], [31, 33, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4], "relations": [[3, 4, 7, 9, "physical", "", false, false], [3, 4, 7, 9, "role", "", false, false], [13, 15, 3, 4, "artifact", "", false, false], [13, 15, 18, 22, "type-of", "", false, false], [13, 15, 31, 33, "related-to", "", true, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["En", "1969", ",", "Victor", "Scheinman", "de", "l'", "universit\u00e9", "de", "Stanford", "a", "invent\u00e9", "le", "bras", "de", "Stanford", ",", "un", "robot", "articul\u00e9", "\u00e0", "6", "axes", ",", "enti\u00e8rement", "\u00e9lectrique", ",", "con\u00e7u", "pour", "permettre", "une", "solution", "de", "bras", "."], "sentence-detokenized": "En 1969, Victor Scheinman de l'universit\u00e9 de Stanford a invent\u00e9 le bras de Stanford, un robot articul\u00e9 \u00e0 6 axes, enti\u00e8rement \u00e9lectrique, con\u00e7u pour permettre une solution de bras.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 15], [16, 25], [26, 28], [29, 31], [31, 41], [42, 44], [45, 53], [54, 55], [56, 63], [64, 66], [67, 71], [72, 74], [75, 83], [83, 84], [85, 87], [88, 93], [94, 102], [103, 104], [105, 106], [107, 111], [111, 112], [113, 124], [125, 135], [135, 136], [137, 142], [143, 147], [148, 157], [158, 161], [162, 170], [171, 173], [174, 178], [178, 179]]}
{"doc_key": "ai-dev-328", "ner": [[8, 8, "product"], [20, 21, "field"], [25, 26, "field"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[8, 8, 20, 21, "related-to", "", false, false], [8, 8, 25, 26, "related-to", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["La", "cr\u00e9ation", "et", "la", "mise", "en", "\u0153uvre", "des", "chatbots", "est", "encore", "un", "domaine", "en", "d\u00e9veloppement", ",", "fortement", "li\u00e9", "\u00e0", "l'", "intelligence", "artificielle", "et", "\u00e0", "l'", "apprentissage", "automatique", ",", "de", "sorte", "que", "les", "solutions", "fournies", ",", "tout", "en", "pr\u00e9sentant", "des", "avantages", "\u00e9vidents", ",", "ont", "des", "limites", "importantes", "en", "termes", "de", "fonctionnalit\u00e9s", "et", "de", "cas", "d'", "utilisation", "."], "sentence-detokenized": "La cr\u00e9ation et la mise en \u0153uvre des chatbots est encore un domaine en d\u00e9veloppement, fortement li\u00e9 \u00e0 l'intelligence artificielle et \u00e0 l'apprentissage automatique, de sorte que les solutions fournies, tout en pr\u00e9sentant des avantages \u00e9vidents, ont des limites importantes en termes de fonctionnalit\u00e9s et de cas d'utilisation.", "token2charspan": [[0, 2], [3, 11], [12, 14], [15, 17], [18, 22], [23, 25], [26, 31], [32, 35], [36, 44], [45, 48], [49, 55], [56, 58], [59, 66], [67, 69], [70, 83], [83, 84], [85, 94], [95, 98], [99, 100], [101, 103], [103, 115], [116, 128], [129, 131], [132, 133], [134, 136], [136, 149], [150, 161], [161, 162], [163, 165], [166, 171], [172, 175], [176, 179], [180, 189], [190, 198], [198, 199], [200, 204], [205, 207], [208, 218], [219, 222], [223, 232], [233, 241], [241, 242], [243, 246], [247, 250], [251, 258], [259, 270], [271, 273], [274, 280], [281, 283], [284, 299], [300, 302], [303, 305], [306, 309], [310, 312], [312, 323], [323, 324]]}
{"doc_key": "ai-dev-329", "ner": [[16, 18, "university"], [10, 13, "product"], [30, 31, "task"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[10, 13, 16, 18, "part-of", "", true, false], [30, 31, 10, 13, "topic", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["En", "ce", "qui", "concerne", "les", "ressources", "disponibles", "gratuitement", ",", "la", "bo\u00eete", "\u00e0", "outils", "Sphinx", "de", "l'", "universit\u00e9", "Carnegie", "Mellon", "est", "un", "bon", "point", "de", "d\u00e9part", "pour", "se", "familiariser", "avec", "la", "reconnaissance", "vocale", "et", "commencer", "\u00e0", "exp\u00e9rimenter", "."], "sentence-detokenized": "En ce qui concerne les ressources disponibles gratuitement, la bo\u00eete \u00e0 outils Sphinx de l'universit\u00e9 Carnegie Mellon est un bon point de d\u00e9part pour se familiariser avec la reconnaissance vocale et commencer \u00e0 exp\u00e9rimenter.", "token2charspan": [[0, 2], [3, 5], [6, 9], [10, 18], [19, 22], [23, 33], [34, 45], [46, 58], [58, 59], [60, 62], [63, 68], [69, 70], [71, 77], [78, 84], [85, 87], [88, 90], [90, 100], [101, 109], [110, 116], [117, 120], [121, 123], [124, 127], [128, 133], [134, 136], [137, 143], [144, 148], [149, 151], [152, 164], [165, 169], [170, 172], [173, 187], [188, 194], [195, 197], [198, 207], [208, 209], [210, 222], [222, 223]]}
{"doc_key": "ai-dev-330", "ner": [[0, 5, "misc"], [12, 22, "misc"], [24, 24, "misc"], [33, 33, "university"], [35, 35, "location"], [38, 38, "country"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[0, 5, 12, 22, "temporal", "", false, false], [24, 24, 12, 22, "named", "", false, false], [24, 24, 35, 35, "physical", "", false, false], [33, 33, 24, 24, "role", "", false, false], [35, 35, 38, 38, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["La", "comp\u00e9tition", "officielle", "de", "la", "RoboCup", "a", "\u00e9t\u00e9", "pr\u00e9c\u00e9d\u00e9e", "par", "le", "premier", "tournoi", "international", "de", "football", "de", "la", "Coupe", "du", "monde", "des", "micro-robots", "(", "MIROSOT", ")", ",", "souvent", "m\u00e9connu", ",", "organis\u00e9", "par", "le", "KAIST", "\u00e0", "Taejon", ",", "en", "Cor\u00e9e", ",", "en", "novembre", "1996", "."], "sentence-detokenized": "La comp\u00e9tition officielle de la RoboCup a \u00e9t\u00e9 pr\u00e9c\u00e9d\u00e9e par le premier tournoi international de football de la Coupe du monde des micro-robots (MIROSOT), souvent m\u00e9connu, organis\u00e9 par le KAIST \u00e0 Taejon, en Cor\u00e9e, en novembre 1996.", "token2charspan": [[0, 2], [3, 14], [15, 25], [26, 28], [29, 31], [32, 39], [40, 41], [42, 45], [46, 54], [55, 58], [59, 61], [62, 69], [70, 77], [78, 91], [92, 94], [95, 103], [104, 106], [107, 109], [110, 115], [116, 118], [119, 124], [125, 128], [129, 141], [142, 143], [143, 150], [150, 151], [151, 152], [153, 160], [161, 168], [168, 169], [170, 178], [179, 182], [183, 185], [186, 191], [192, 193], [194, 200], [200, 201], [202, 204], [205, 210], [210, 211], [212, 214], [215, 223], [224, 228], [228, 229]]}
{"doc_key": "ai-dev-331", "ner": [[4, 7, "metrics"], [25, 27, "misc"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["En", "plus", "de", "la", "perte", "standard", "de", "charni\u00e8re", "math", "(", "1-yf", "(", "x", ")", ")", "_", "+", "/", "math", "pour", "les", "donn\u00e9es", "\u00e9tiquet\u00e9es", ",", "une", "fonction", "de", "perte", "math", "(", "-1", "|", "f", "(", "x", ")", "|", ")", "_", "+", "/", "math", "est", "introduite", "sur", "les", "donn\u00e9es", "non", "\u00e9tiquet\u00e9es", "en", "laissant", "mathy", "=", "\\", "nom", "d'", "op\u00e9rateur", "{sign}", "{f", "(", "x)}", "/", "math", "."], "sentence-detokenized": "En plus de la perte standard de charni\u00e8re math (1-yf (x)) _ + / math pour les donn\u00e9es \u00e9tiquet\u00e9es, une fonction de perte math (-1 | f (x) |) _ + / math est introduite sur les donn\u00e9es non \u00e9tiquet\u00e9es en laissant mathy =\\ nom d'op\u00e9rateur {sign} {f (x)} / math.", "token2charspan": [[0, 2], [3, 7], [8, 10], [11, 13], [14, 19], [20, 28], [29, 31], [32, 41], [42, 46], [47, 48], [48, 52], [53, 54], [54, 55], [55, 56], [56, 57], [58, 59], [60, 61], [62, 63], [64, 68], [69, 73], [74, 77], [78, 85], [86, 96], [96, 97], [98, 101], [102, 110], [111, 113], [114, 119], [120, 124], [125, 126], [126, 128], [129, 130], [131, 132], [133, 134], [134, 135], [135, 136], [137, 138], [138, 139], [140, 141], [142, 143], [144, 145], [146, 150], [151, 154], [155, 165], [166, 169], [170, 173], [174, 181], [182, 185], [186, 196], [197, 199], [200, 208], [209, 214], [215, 216], [216, 217], [218, 221], [222, 224], [224, 233], [234, 240], [241, 243], [244, 245], [245, 248], [249, 250], [251, 255], [255, 256]]}
{"doc_key": "ai-dev-332", "ner": [[3, 3, "misc"], [9, 11, "metrics"]], "ner_mapping_to_source": [0, 1], "relations": [[3, 3, 9, 11, "related-to", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["En", "particulier", ",", "RLS", "est", "con\u00e7u", "pour", "minimiser", "l'", "erreur", "quadratique", "moyenne", "entre", "les", "valeurs", "pr\u00e9dites", "et", "les", "\u00e9tiquettes", "VRAIES", ",", "sous", "r\u00e9serve", "de", "r\u00e9gularisation", "."], "sentence-detokenized": "En particulier, RLS est con\u00e7u pour minimiser l'erreur quadratique moyenne entre les valeurs pr\u00e9dites et les \u00e9tiquettes VRAIES, sous r\u00e9serve de r\u00e9gularisation.", "token2charspan": [[0, 2], [3, 14], [14, 15], [16, 19], [20, 23], [24, 29], [30, 34], [35, 44], [45, 47], [47, 53], [54, 65], [66, 73], [74, 79], [80, 83], [84, 91], [92, 100], [101, 103], [104, 107], [108, 118], [119, 125], [125, 126], [127, 131], [132, 139], [140, 142], [143, 157], [157, 158]]}
{"doc_key": "ai-dev-333", "ner": [[5, 9, "algorithm"], [12, 14, "algorithm"]], "ner_mapping_to_source": [0, 1], "relations": [[5, 9, 12, 14, "related-to", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Essentiellement", ",", "cela", "combine", "l'", "estimation", "du", "maximum", "de", "vraisemblance", "avec", "une", "proc\u00e9dure", "de", "r\u00e9gularisation", "qui", "favorise", "les", "mod\u00e8les", "plus", "simples", "par", "rapport", "aux", "mod\u00e8les", "plus", "complexes", "."], "sentence-detokenized": "Essentiellement, cela combine l'estimation du maximum de vraisemblance avec une proc\u00e9dure de r\u00e9gularisation qui favorise les mod\u00e8les plus simples par rapport aux mod\u00e8les plus complexes.", "token2charspan": [[0, 15], [15, 16], [17, 21], [22, 29], [30, 32], [32, 42], [43, 45], [46, 53], [54, 56], [57, 70], [71, 75], [76, 79], [80, 89], [90, 92], [93, 107], [108, 111], [112, 120], [121, 124], [125, 132], [133, 137], [138, 145], [146, 149], [150, 157], [158, 161], [162, 169], [170, 174], [175, 184], [184, 185]]}
{"doc_key": "ai-dev-334", "ner": [[1, 4, "metrics"], [8, 8, "metrics"], [10, 10, "metrics"], [12, 14, "misc"], [17, 19, "misc"], [35, 38, "algorithm"], [41, 44, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6], "relations": [[8, 8, 1, 4, "named", "", false, false], [10, 10, 1, 4, "named", "", false, false], [12, 14, 17, 19, "related-to", "", false, false], [12, 14, 35, 38, "related-to", "ratio", false, false], [35, 38, 41, 44, "related-to", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Le", "taux", "de", "vrais", "positifs", "est", "\u00e9galement", "appel\u00e9", "sensibilit\u00e9", ",", "rappel", "ou", "probabilit\u00e9", "de", "d\u00e9tection", "math\u00e9matique", "au", "seuil", "de", "discrimination", ")", "de", "la", "probabilit\u00e9", "de", "d\u00e9tection", "sur", "l'", "axe", "des", "y", "par", "rapport", "\u00e0", "la", "fonction", "de", "distribution", "cumulative", "de", "la", "probabilit\u00e9", "de", "fausse", "alerte", "sur", "l'", "axe", "des", "x", "."], "sentence-detokenized": "Le taux de vrais positifs est \u00e9galement appel\u00e9 sensibilit\u00e9, rappel ou probabilit\u00e9 de d\u00e9tection math\u00e9matique au seuil de discrimination) de la probabilit\u00e9 de d\u00e9tection sur l'axe des y par rapport \u00e0 la fonction de distribution cumulative de la probabilit\u00e9 de fausse alerte sur l'axe des x.", "token2charspan": [[0, 2], [3, 7], [8, 10], [11, 16], [17, 25], [26, 29], [30, 39], [40, 46], [47, 58], [58, 59], [60, 66], [67, 69], [70, 81], [82, 84], [85, 94], [95, 107], [108, 110], [111, 116], [117, 119], [120, 134], [134, 135], [136, 138], [139, 141], [142, 153], [154, 156], [157, 166], [167, 170], [171, 173], [173, 176], [177, 180], [181, 182], [183, 186], [187, 194], [195, 196], [197, 199], [200, 208], [209, 211], [212, 224], [225, 235], [236, 238], [239, 241], [242, 253], [254, 256], [257, 263], [264, 270], [271, 274], [275, 277], [277, 280], [281, 284], [285, 286], [286, 287]]}
{"doc_key": "ai-dev-335", "ner": [[1, 1, "misc"], [3, 3, "product"]], "ner_mapping_to_source": [0, 1], "relations": [[3, 3, 1, 1, "general-affiliation", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["En", "anglais", ",", "WordNet", "est", "un", "exemple", "de", "r\u00e9seau", "s\u00e9mantique", "."], "sentence-detokenized": "En anglais, WordNet est un exemple de r\u00e9seau s\u00e9mantique.", "token2charspan": [[0, 2], [3, 10], [10, 11], [12, 19], [20, 23], [24, 26], [27, 34], [35, 37], [38, 44], [45, 55], [55, 56]]}
{"doc_key": "ai-dev-336", "ner": [[5, 8, "product"], [12, 14, "product"], [33, 36, "misc"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[33, 36, 5, 8, "usage", "", false, false], [33, 36, 12, 14, "usage", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["L'", "utilisation", "prolong\u00e9e", "d'", "un", "logiciel", "de", "reconnaissance", "vocale", "associ\u00e9", "\u00e0", "un", "traitement", "de", "texte", "a", "montr\u00e9", "des", "avantages", "pour", "le", "renforcement", "de", "la", "m\u00e9moire", "\u00e0", "court", "terme", "chez", "les", "patients", "atteints", "de", "MAV", "c\u00e9r\u00e9brale", "qui", "ont", "\u00e9t\u00e9", "trait\u00e9s", "par", "r\u00e9section", "."], "sentence-detokenized": "L'utilisation prolong\u00e9e d'un logiciel de reconnaissance vocale associ\u00e9 \u00e0 un traitement de texte a montr\u00e9 des avantages pour le renforcement de la m\u00e9moire \u00e0 court terme chez les patients atteints de MAV c\u00e9r\u00e9brale qui ont \u00e9t\u00e9 trait\u00e9s par r\u00e9section.", "token2charspan": [[0, 2], [2, 13], [14, 23], [24, 26], [26, 28], [29, 37], [38, 40], [41, 55], [56, 62], [63, 70], [71, 72], [73, 75], [76, 86], [87, 89], [90, 95], [96, 97], [98, 104], [105, 108], [109, 118], [119, 123], [124, 126], [127, 139], [140, 142], [143, 145], [146, 153], [154, 155], [156, 161], [162, 167], [168, 172], [173, 176], [177, 185], [186, 194], [195, 197], [198, 201], [202, 211], [212, 215], [216, 219], [220, 223], [224, 231], [232, 235], [236, 245], [245, 246]]}
{"doc_key": "ai-dev-337", "ner": [[6, 7, "researcher"], [9, 10, "researcher"], [12, 13, "researcher"]], "ner_mapping_to_source": [0, 1, 2], "relations": [], "relations_mapping_to_source": [], "sentence": ["Ses", "r\u00e9dacteurs", "en", "chef", "fondateurs", "\u00e9taient", "Ron", "Sun", ",", "Vasant", "Honavar", "et", "Gregg", "Oden", "(", "de", "1999", "\u00e0", "2014", ")", "."], "sentence-detokenized": "Ses r\u00e9dacteurs en chef fondateurs \u00e9taient Ron Sun, Vasant Honavar et Gregg Oden (de 1999 \u00e0 2014).", "token2charspan": [[0, 3], [4, 14], [15, 17], [18, 22], [23, 33], [34, 41], [42, 45], [46, 49], [49, 50], [51, 57], [58, 65], [66, 68], [69, 74], [75, 79], [80, 81], [81, 83], [84, 88], [89, 90], [91, 95], [95, 96], [96, 97]]}
{"doc_key": "ai-dev-338", "ner": [[10, 12, "product"], [17, 18, "misc"]], "ner_mapping_to_source": [0, 1], "relations": [[10, 12, 17, 18, "opposite", "", false, false]], "relations_mapping_to_source": [0], "sentence": ["Leur", "distinction", "\"", "parall\u00e8le", "\"", ",", "par", "opposition", "\u00e0", "un", "manipulateur", "en", "s\u00e9rie", ",", "est", "que", "l'", "effecteur", "final", "(", "ou", "\"", "main", "\"", ")", "de", "cette", "liaison", "(", "ou", "\"", "bras", "\"", ")", "est", "directement", "reli\u00e9", "\u00e0", "sa", "base", "par", "un", "certain", "nombre", "(", "g\u00e9n\u00e9ralement", "trois", "ou", "six", ")", "de", "liaisons", "s\u00e9par\u00e9es", "et", "ind\u00e9pendantes", "fonctionnant", "simultan\u00e9ment", "."], "sentence-detokenized": "Leur distinction \"parall\u00e8le\", par opposition \u00e0 un manipulateur en s\u00e9rie, est que l'effecteur final (ou \"main\") de cette liaison (ou \"bras\") est directement reli\u00e9 \u00e0 sa base par un certain nombre (g\u00e9n\u00e9ralement trois ou six) de liaisons s\u00e9par\u00e9es et ind\u00e9pendantes fonctionnant simultan\u00e9ment.", "token2charspan": [[0, 4], [5, 16], [17, 18], [18, 27], [27, 28], [28, 29], [30, 33], [34, 44], [45, 46], [47, 49], [50, 62], [63, 65], [66, 71], [71, 72], [73, 76], [77, 80], [81, 83], [83, 92], [93, 98], [99, 100], [100, 102], [103, 104], [104, 108], [108, 109], [109, 110], [111, 113], [114, 119], [120, 127], [128, 129], [129, 131], [132, 133], [133, 137], [137, 138], [138, 139], [140, 143], [144, 155], [156, 161], [162, 163], [164, 166], [167, 171], [172, 175], [176, 178], [179, 186], [187, 193], [194, 195], [195, 207], [208, 213], [214, 216], [217, 220], [220, 221], [222, 224], [225, 233], [234, 242], [243, 245], [246, 259], [260, 272], [273, 286], [286, 287]]}
{"doc_key": "ai-dev-339", "ner": [[7, 8, "researcher"], [21, 22, "researcher"], [23, 25, "researcher"], [27, 28, "researcher"], [30, 31, "researcher"], [33, 34, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [], "relations_mapping_to_source": [], "sentence": ["Son", "directeur", "de", "th\u00e8se", "\u00e9tait", "le", "professeur", "Cordell", "Green", ",", "et", "son", "comit\u00e9", "de", "th\u00e8se", "/", "oral", "\u00e9tait", "compos\u00e9", "des", "professeurs", "Edward", "Feigenbaum", ",", "Joshua", "Lederberg", ",", "Paul", "Cohen", ",", "Allen", "Newell", ",", "Herbert", "Simon", "."], "sentence-detokenized": "Son directeur de th\u00e8se \u00e9tait le professeur Cordell Green, et son comit\u00e9 de th\u00e8se/oral \u00e9tait compos\u00e9 des professeurs Edward Feigenbaum, Joshua Lederberg, Paul Cohen, Allen Newell, Herbert Simon.", "token2charspan": [[0, 3], [4, 13], [14, 16], [17, 22], [23, 28], [29, 31], [32, 42], [43, 50], [51, 56], [56, 57], [58, 60], [61, 64], [65, 71], [72, 74], [75, 80], [80, 81], [81, 85], [86, 91], [92, 99], [100, 103], [104, 115], [116, 122], [123, 133], [133, 134], [135, 141], [142, 151], [151, 152], [153, 157], [158, 163], [163, 164], [165, 170], [171, 177], [177, 178], [179, 186], [187, 192], [192, 193]]}
{"doc_key": "ai-dev-340", "ner": [[4, 6, "metrics"], [9, 12, "metrics"], [15, 17, "metrics"], [20, 22, "metrics"], [25, 28, "metrics"], [31, 33, "metrics"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [], "relations_mapping_to_source": [], "sentence": ["Ces", "fonctions", "comprennent", "l'", "erreur", "quadratique", "moyenne", ",", "l'", "erreur", "quadratique", "moyenne", "racine", ",", "l'", "erreur", "absolue", "moyenne", ",", "l'", "erreur", "quadratique", "relative", ",", "l'", "erreur", "quadratique", "relative", "racine", ",", "l'", "erreur", "absolue", "relative", ",", "et", "autres", "."], "sentence-detokenized": "Ces fonctions comprennent l'erreur quadratique moyenne, l'erreur quadratique moyenne racine, l'erreur absolue moyenne, l'erreur quadratique relative, l'erreur quadratique relative racine, l'erreur absolue relative, et autres.", "token2charspan": [[0, 3], [4, 13], [14, 25], [26, 28], [28, 34], [35, 46], [47, 54], [54, 55], [56, 58], [58, 64], [65, 76], [77, 84], [85, 91], [91, 92], [93, 95], [95, 101], [102, 109], [110, 117], [117, 118], [119, 121], [121, 127], [128, 139], [140, 148], [148, 149], [150, 152], [152, 158], [159, 170], [171, 179], [180, 186], [186, 187], [188, 190], [190, 196], [197, 204], [205, 213], [213, 214], [215, 217], [218, 224], [224, 225]]}
{"doc_key": "ai-dev-341", "ner": [[5, 5, "programlang"], [7, 7, "programlang"], [9, 9, "product"], [11, 11, "programlang"]], "ner_mapping_to_source": [0, 1, 2, 3], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "existe", "des", "liaisons", "en", "Python", ",", "Java", "et", "MATLAB", "/", "OCTAVE", "."], "sentence-detokenized": "Il existe des liaisons en Python, Java et MATLAB / OCTAVE.", "token2charspan": [[0, 2], [3, 9], [10, 13], [14, 22], [23, 25], [26, 32], [32, 33], [34, 38], [39, 41], [42, 48], [49, 50], [51, 57], [57, 58]]}
{"doc_key": "ai-dev-342", "ner": [[3, 3, "product"]], "ner_mapping_to_source": [0], "relations": [], "relations_mapping_to_source": [], "sentence": ["Une", "impl\u00e9mentation", "dans", "MATLAB", "peut", "\u00eatre", "trouv\u00e9e", "sur", "le", "site", "."], "sentence-detokenized": "Une impl\u00e9mentation dans MATLAB peut \u00eatre trouv\u00e9e sur le site.", "token2charspan": [[0, 3], [4, 18], [19, 23], [24, 30], [31, 35], [36, 40], [41, 48], [49, 52], [53, 55], [56, 60], [60, 61]]}
{"doc_key": "ai-dev-343", "ner": [[0, 1, "researcher"], [10, 11, "field"], [14, 15, "researcher"], [17, 18, "researcher"], [20, 21, "researcher"], [23, 25, "researcher"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[10, 11, 0, 1, "origin", "", false, false], [10, 11, 14, 15, "origin", "", false, false], [10, 11, 17, 18, "origin", "", false, false], [10, 11, 20, 21, "origin", "", false, false], [10, 11, 23, 25, "origin", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["John", "McCarthy", "est", "l'", "un", "des", "p\u00e8res", "fondateurs", "de", "l'", "intelligence", "artificielle", ",", "avec", "Alan", "Turing", ",", "Marvin", "Minsky", ",", "Allen", "Newell", "et", "Herbert", "A.", "Simon", "."], "sentence-detokenized": "John McCarthy est l'un des p\u00e8res fondateurs de l'intelligence artificielle, avec Alan Turing, Marvin Minsky, Allen Newell et Herbert A. Simon.", "token2charspan": [[0, 4], [5, 13], [14, 17], [18, 20], [20, 22], [23, 26], [27, 32], [33, 43], [44, 46], [47, 49], [49, 61], [62, 74], [74, 75], [76, 80], [81, 85], [86, 92], [92, 93], [94, 100], [101, 107], [107, 108], [109, 114], [115, 121], [122, 124], [125, 132], [133, 135], [136, 141], [141, 142]]}
{"doc_key": "ai-dev-344", "ner": [[10, 12, "product"], [20, 21, "misc"]], "ner_mapping_to_source": [0, 1], "relations": [], "relations_mapping_to_source": [], "sentence": ["Un", "manipulateur", "parall\u00e8le", "est", "un", "syst\u00e8me", "m\u00e9canique", "qui", "utilise", "plusieurs", "manipulateurs", "en", "s\u00e9rie", "pour", "supporter", "une", "seule", "plate-forme", ",", "ou", "effecteur", "final", "."], "sentence-detokenized": "Un manipulateur parall\u00e8le est un syst\u00e8me m\u00e9canique qui utilise plusieurs manipulateurs en s\u00e9rie pour supporter une seule plate-forme, ou effecteur final.", "token2charspan": [[0, 2], [3, 15], [16, 25], [26, 29], [30, 32], [33, 40], [41, 50], [51, 54], [55, 62], [63, 72], [73, 86], [87, 89], [90, 95], [96, 100], [101, 110], [111, 114], [115, 120], [121, 132], [132, 133], [134, 136], [137, 146], [147, 152], [152, 153]]}
{"doc_key": "ai-dev-345", "ner": [[0, 0, "product"], [5, 7, "task"], [9, 9, "product"], [11, 15, "product"], [26, 26, "misc"], [29, 29, "misc"], [32, 34, "misc"], [37, 38, "task"], [41, 46, "product"], [49, 50, "product"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9], "relations": [[9, 9, 0, 0, "part-of", "", false, false], [9, 9, 5, 7, "type-of", "", false, false], [11, 15, 9, 9, "named", "", false, false], [26, 26, 9, 9, "part-of", "", false, false], [29, 29, 9, 9, "part-of", "", false, false], [32, 34, 9, 9, "part-of", "", false, false], [37, 38, 9, 9, "part-of", "", false, false], [41, 46, 9, 9, "part-of", "", false, false], [49, 50, 9, 9, "part-of", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4, 5, 6, 7, 8], "sentence": ["GATE", "comprend", "un", "syst\u00e8me", "d'", "extraction", "d'", "informations", "appel\u00e9", "ANNIE", "(", "A", "Nearly-New", "Information", "Extraction", "System", ")", ",", "qui", "est", "un", "ensemble", "de", "modules", "comprenant", "un", "tokenizer", ",", "un", "gazetteer", ",", "un", "s\u00e9parateur", "de", "phrases", ",", "un", "tagging", "Part-of-speech", ",", "un", "transducteur", "de", "reconnaissance", "d'", "entit\u00e9s", "nomm\u00e9es", "et", "un", "tagger", "coreference", "."], "sentence-detokenized": "GATE comprend un syst\u00e8me d'extraction d'informations appel\u00e9 ANNIE (A Nearly-New Information Extraction System), qui est un ensemble de modules comprenant un tokenizer, un gazetteer, un s\u00e9parateur de phrases, un tagging Part-of-speech, un transducteur de reconnaissance d'entit\u00e9s nomm\u00e9es et un tagger coreference.", "token2charspan": [[0, 4], [5, 13], [14, 16], [17, 24], [25, 27], [27, 37], [38, 40], [40, 52], [53, 59], [60, 65], [66, 67], [67, 68], [69, 79], [80, 91], [92, 102], [103, 109], [109, 110], [110, 111], [112, 115], [116, 119], [120, 122], [123, 131], [132, 134], [135, 142], [143, 153], [154, 156], [157, 166], [166, 167], [168, 170], [171, 180], [180, 181], [182, 184], [185, 195], [196, 198], [199, 206], [206, 207], [208, 210], [211, 218], [219, 233], [233, 234], [235, 237], [238, 250], [251, 253], [254, 268], [269, 271], [271, 278], [279, 286], [287, 289], [290, 292], [293, 299], [300, 311], [311, 312]]}
{"doc_key": "ai-dev-346", "ner": [[5, 9, "university"], [19, 19, "country"], [26, 29, "person"]], "ner_mapping_to_source": [0, 1, 2], "relations": [], "relations_mapping_to_source": [], "sentence": ["Il", "est", "dipl\u00f4m\u00e9", "de", "l'", "universit\u00e9", "d'", "\u00c9tat", "de", "Moscou", "et", "en", "novembre", "1978", ",", "il", "part", "pour", "les", "\u00c9tats-Unis", "gr\u00e2ce", "\u00e0", "l'", "intervention", "personnelle", "du", "s\u00e9nateur", "Edward", "M.", "Kennedy", "."], "sentence-detokenized": "Il est dipl\u00f4m\u00e9 de l'universit\u00e9 d'\u00c9tat de Moscou et en novembre 1978, il part pour les \u00c9tats-Unis gr\u00e2ce \u00e0 l'intervention personnelle du s\u00e9nateur Edward M. Kennedy .", "token2charspan": [[0, 2], [3, 6], [7, 14], [15, 17], [18, 20], [20, 30], [31, 33], [33, 37], [38, 40], [41, 47], [48, 50], [51, 53], [54, 62], [63, 67], [67, 68], [69, 71], [72, 76], [77, 81], [82, 85], [86, 96], [97, 102], [103, 104], [105, 107], [107, 119], [120, 131], [132, 134], [135, 143], [144, 150], [151, 153], [154, 161], [162, 163]]}
{"doc_key": "ai-dev-347", "ner": [[4, 6, "organisation"], [10, 16, "misc"], [24, 24, "field"]], "ner_mapping_to_source": [0, 1, 2], "relations": [[4, 6, 10, 16, "win-defeat", "", false, false], [10, 16, 24, 24, "topic", "", false, false]], "relations_mapping_to_source": [0, 1], "sentence": ["En", "2017", ",", "l'", "\u00e9quipe", "DeepMind", "AlphaGo", "a", "re\u00e7u", "la", "premi\u00e8re", "m\u00e9daille", "Marvin", "Minsky", "de", "l'", "IJCAI", "pour", "ses", "r\u00e9alisations", "exceptionnelles", "en", "mati\u00e8re", "d'", "IA", "."], "sentence-detokenized": "En 2017, l'\u00e9quipe DeepMind AlphaGo a re\u00e7u la premi\u00e8re m\u00e9daille Marvin Minsky de l'IJCAI pour ses r\u00e9alisations exceptionnelles en mati\u00e8re d'IA.", "token2charspan": [[0, 2], [3, 7], [7, 8], [9, 11], [11, 17], [18, 26], [27, 34], [35, 36], [37, 41], [42, 44], [45, 53], [54, 62], [63, 69], [70, 76], [77, 79], [80, 82], [82, 87], [88, 92], [93, 96], [97, 109], [110, 125], [126, 128], [129, 136], [137, 139], [139, 141], [141, 142]]}
{"doc_key": "ai-dev-348", "ner": [[0, 3, "misc"], [8, 8, "misc"], [15, 15, "misc"], [27, 33, "misc"], [34, 34, "misc"], [40, 40, "misc"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[0, 3, 8, 8, "related-to", "is_recorded_by", false, false], [8, 8, 15, 15, "cause-effect", "", false, false], [8, 8, 15, 15, "physical", "", false, false], [8, 8, 27, 33, "physical", "", false, false], [8, 8, 40, 40, "physical", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["La", "propagation", "anormale", "est", "\u00e9galement", "enregistr\u00e9e", "par", "les", "troposcatters", "qui", "provoquent", "des", "irr\u00e9gularit\u00e9s", "dans", "la", "troposph\u00e8re", ",", "la", "diffusion", "due", "aux", "m\u00e9t\u00e9ores", ",", "la", "r\u00e9fraction", "dans", "les", "r\u00e9gions", "et", "les", "couches", "ionis\u00e9es", "de", "l'", "ionosph\u00e8re", "et", "la", "r\u00e9flexion", "sur", "l'", "ionosph\u00e8re", "."], "sentence-detokenized": "La propagation anormale est \u00e9galement enregistr\u00e9e par les troposcatters qui provoquent des irr\u00e9gularit\u00e9s dans la troposph\u00e8re, la diffusion due aux m\u00e9t\u00e9ores, la r\u00e9fraction dans les r\u00e9gions et les couches ionis\u00e9es de l'ionosph\u00e8re et la r\u00e9flexion sur l'ionosph\u00e8re.", "token2charspan": [[0, 2], [3, 14], [15, 23], [24, 27], [28, 37], [38, 49], [50, 53], [54, 57], [58, 71], [72, 75], [76, 86], [87, 90], [91, 104], [105, 109], [110, 112], [113, 124], [124, 125], [126, 128], [129, 138], [139, 142], [143, 146], [147, 155], [155, 156], [157, 159], [160, 170], [171, 175], [176, 179], [180, 187], [188, 190], [191, 194], [195, 202], [203, 211], [212, 214], [215, 217], [217, 227], [228, 230], [231, 233], [234, 243], [244, 247], [248, 250], [250, 260], [260, 261]]}
{"doc_key": "ai-dev-349", "ner": [[0, 4, "field"], [6, 6, "field"], [13, 13, "field"], [17, 17, "field"], [21, 24, "field"], [28, 29, "field"]], "ner_mapping_to_source": [0, 1, 2, 3, 4, 5], "relations": [[0, 4, 13, 13, "part-of", "", false, false], [0, 4, 17, 17, "part-of", "", false, false], [0, 4, 21, 24, "part-of", "", false, false], [0, 4, 28, 29, "part-of", "", false, false], [6, 6, 0, 4, "named", "", false, false]], "relations_mapping_to_source": [0, 1, 2, 3, 4], "sentence": ["Le", "traitement", "du", "langage", "naturel", "(", "NLP", ")", "est", "un", "sous-domaine", "de", "la", "linguistique", ",", "de", "l'", "informatique", ",", "de", "l'", "ing\u00e9nierie", "de", "l'", "information", "et", "de", "l'", "intelligence", "artificielle", "qui", "s'", "int\u00e9resse", "aux", "interactions", "entre", "les", "ordinateurs", "et", "les", "langues", "(", "naturelles", ")", "humaines", ",", "en", "particulier", "\u00e0", "la", "mani\u00e8re", "de", "programmer", "les", "ordinateurs", "pour", "traiter", "et", "analyser", "de", "grandes", "quantit\u00e9s", "de", "donn\u00e9es", "en", "langage", "naturel", "."], "sentence-detokenized": "Le traitement du langage naturel (NLP) est un sous-domaine de la linguistique, de l'informatique, de l'ing\u00e9nierie de l'information et de l'intelligence artificielle qui s'int\u00e9resse aux interactions entre les ordinateurs et les langues (naturelles) humaines, en particulier \u00e0 la mani\u00e8re de programmer les ordinateurs pour traiter et analyser de grandes quantit\u00e9s de donn\u00e9es en langage naturel.", "token2charspan": [[0, 2], [3, 13], [14, 16], [17, 24], [25, 32], [33, 34], [34, 37], [37, 38], [39, 42], [43, 45], [46, 58], [59, 61], [62, 64], [65, 77], [77, 78], [79, 81], [82, 84], [84, 96], [96, 97], [98, 100], [101, 103], [103, 113], [114, 116], [117, 119], [119, 130], [131, 133], [134, 136], [137, 139], [139, 151], [152, 164], [165, 168], [169, 171], [171, 180], [181, 184], [185, 197], [198, 203], [204, 207], [208, 219], [220, 222], [223, 226], [227, 234], [235, 236], [236, 246], [246, 247], [248, 256], [256, 257], [258, 260], [261, 272], [273, 274], [275, 277], [278, 285], [286, 288], [289, 299], [300, 303], [304, 315], [316, 320], [321, 328], [329, 331], [332, 340], [341, 343], [344, 351], [352, 361], [362, 364], [365, 372], [373, 375], [376, 383], [384, 391], [391, 392]]}
{"doc_key": "ai-dev-350", "ner": [[14, 15, "organisation"], [18, 19, "organisation"], [21, 23, "organisation"]], "ner_mapping_to_source": [0, 1, 2], "relations": [], "relations_mapping_to_source": [], "sentence": ["Parmi", "les", "autres", "groupes", "de", "jeunes", "actifs", "dans", "le", "domaine", "du", "climat", ",", "citons", "Extinction", "Rebellion", ",", "le", "Sunrise", "Movement", ",", "SustainUS", "et", "d'", "autres", "groupes", "travaillant", "aux", "niveaux", "transnational", "et", "local", "."], "sentence-detokenized": "Parmi les autres groupes de jeunes actifs dans le domaine du climat, citons Extinction Rebellion, le Sunrise Movement, SustainUS et d'autres groupes travaillant aux niveaux transnational et local.", "token2charspan": [[0, 5], [6, 9], [10, 16], [17, 24], [25, 27], [28, 34], [35, 41], [42, 46], [47, 49], [50, 57], [58, 60], [61, 67], [67, 68], [69, 75], [76, 86], [87, 96], [96, 97], [98, 100], [101, 108], [109, 117], [117, 118], [119, 128], [129, 131], [132, 134], [134, 140], [141, 148], [149, 160], [161, 164], [165, 172], [173, 186], [187, 189], [190, 195], [195, 196]]}