| { |
| "paper_id": "P93-1004", |
| "header": { |
| "generated_with": "S2ORC 1.0.0", |
| "date_generated": "2023-01-19T08:52:04.996977Z" |
| }, |
| "title": "", |
| "authors": [ |
| { |
| "first": "Yuji", |
| "middle": [], |
| "last": "Matsulnoto", |
| "suffix": "", |
| "affiliation": { |
| "laboratory": "", |
| "institution": "Nara Takayanaa-cho", |
| "location": { |
| "addrLine": "Ikoma-shi, Na.ra", |
| "postCode": "630-01", |
| "country": "Japan" |
| } |
| }, |
| "email": "" |
| }, |
| { |
| "first": "Hiroyuki", |
| "middle": [], |
| "last": "Ishimoto", |
| "suffix": "", |
| "affiliation": { |
| "laboratory": "", |
| "institution": "Kyoto University", |
| "location": { |
| "addrLine": "Sakyo-ku", |
| "postCode": "606", |
| "settlement": "Kyoto", |
| "country": "Japan" |
| } |
| }, |
| "email": "ishimoto@pine.kuee.kyoto-u.ac.jp" |
| }, |
| { |
| "first": "Takehito", |
| "middle": [], |
| "last": "Utsuro", |
| "suffix": "", |
| "affiliation": { |
| "laboratory": "", |
| "institution": "Kyoto University", |
| "location": { |
| "addrLine": "Sakyo-ku", |
| "postCode": "606", |
| "settlement": "Kyoto", |
| "country": "Japan" |
| } |
| }, |
| "email": "utsuro@pine.kuee.kyoto-u.ac.jp" |
| } |
| ], |
| "year": "", |
| "venue": null, |
| "identifiers": {}, |
| "abstract": "This paper describes a method for finding strucrural matching between parallel sentences of two languages, (such as Japanese and English). Parallel sentences are analyzed based on unification grammars, and structural matching is performed by making use of a similarity measure of word pairs in the two languages. Syntactic ambiguities are resolved simultaneously in the matching process. The results serve as a. useful source for extracting linguistic a.nd lexical knowledge.", |
| "pdf_parse": { |
| "paper_id": "P93-1004", |
| "_pdf_hash": "", |
| "abstract": [ |
| { |
| "text": "This paper describes a method for finding strucrural matching between parallel sentences of two languages, (such as Japanese and English). Parallel sentences are analyzed based on unification grammars, and structural matching is performed by making use of a similarity measure of word pairs in the two languages. Syntactic ambiguities are resolved simultaneously in the matching process. The results serve as a. useful source for extracting linguistic a.nd lexical knowledge.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Abstract", |
| "sec_num": null |
| } |
| ], |
| "body_text": [ |
| { |
| "text": "Bilingual (or parallel) texts are useful resources for acquisition of linguistic knowledge as well as for applications such as machine translation. Intensive research has been done for aligning bilingual texts at the sentence level using statistical teclmiques by measuring sentence lengths in words or in characters (Brown 91) , (Gale 91a) . Those works are quite successful in that far more than 90% of sentences in bilingual corpora, are a.ligned correctly.", |
| "cite_spans": [ |
| { |
| "start": 317, |
| "end": 327, |
| "text": "(Brown 91)", |
| "ref_id": null |
| }, |
| { |
| "start": 330, |
| "end": 340, |
| "text": "(Gale 91a)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "INTRODUCTION", |
| "sec_num": null |
| }, |
| { |
| "text": "Although such parallel texts are shown to be useful in real applications such as machine translation (Brown 90) and word sense disambiguatioll (Daga.n 91), structured bilingual sentences are undoubtedly more informative and important for filture natural language researches. Structured bilingual or multilingual corpora, serve a.s richer sources for extracting linguistic knowledge (Kaji 92) , (Klavans 90) , (Sadler 91) , (Utsuro 92) .", |
| "cite_spans": [ |
| { |
| "start": 101, |
| "end": 111, |
| "text": "(Brown 90)", |
| "ref_id": null |
| }, |
| { |
| "start": 382, |
| "end": 391, |
| "text": "(Kaji 92)", |
| "ref_id": null |
| }, |
| { |
| "start": 394, |
| "end": 406, |
| "text": "(Klavans 90)", |
| "ref_id": null |
| }, |
| { |
| "start": 409, |
| "end": 420, |
| "text": "(Sadler 91)", |
| "ref_id": null |
| }, |
| { |
| "start": 423, |
| "end": 434, |
| "text": "(Utsuro 92)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "INTRODUCTION", |
| "sec_num": null |
| }, |
| { |
| "text": "Phrase level or word level alignment has also been done by several researchers. The Textual Knowledge Bank Project (Sadler 91) is building lnonolingual and multilingual text bases structured by linking the elements with grammatical (dependency), referential, and bilingual relations. (Karl 92) reports a method to obtain phrase level correspondence of parallel texts by coupling phrases of two languages obtained in CKY parsing processes. This paper presents another method to obtain structural matching of bilingual texts. Sentences in both languages are parsed to produce (disjunctive) feature structures, from which dependency structures are extracted. Ambiguities are represented as disjunction. Then, the two structures are matched to establish a one-to-one correspondence between their substructures. The result of the match is obtained as a set of pairs of minimal corresponding substructures of the dependency structures. Examples of the results are shown in Figures 1, 2 and 3 . A dependency structure is represented as a tree, in which ambiguity is specified by a disjunctive node (OR. node). Circles in the figure show substructures and bidirectional arrows show corresponding substructures.", |
| "cite_spans": [ |
| { |
| "start": 115, |
| "end": 126, |
| "text": "(Sadler 91)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [ |
| { |
| "start": 967, |
| "end": 985, |
| "text": "Figures 1, 2 and 3", |
| "ref_id": "FIGREF0" |
| } |
| ], |
| "eq_spans": [], |
| "section": "INTRODUCTION", |
| "sec_num": null |
| }, |
| { |
| "text": "Our technique and the results are different from those of other lnethods mentioned above. (Kaji 92) identifies corresponding phrases and ahns at producing tra.nslation templates by abstracting those corresponding phrases. In the Bilingua.l Knowledge Bank (Sadler 91) , the correspondence is shown by links between words in two sentences, equating two whole subtrees headed by the words. We prefer the Ininimal substructure correspondence and the relationship between substructures. Such a minimal substructure stands for the minimal meaningful component in the sentence, which we believe is very useful for our target application of extracting lexical knowledge fi'om bilingual corpora.", |
| "cite_spans": [ |
| { |
| "start": 90, |
| "end": 99, |
| "text": "(Kaji 92)", |
| "ref_id": null |
| }, |
| { |
| "start": 255, |
| "end": 266, |
| "text": "(Sadler 91)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "INTRODUCTION", |
| "sec_num": null |
| }, |
| { |
| "text": "Although the structural matching method shown in this paper is language independent, we deal with parallel texts of Japanese a.nd English. We assume that a.lignment at the sentence level is already preprocessed manually or by other methods such as those in (Brown 91) , (Gale 91a) . Throughout this paper, we assume to match simple sentences. 1", |
| "cite_spans": [ |
| { |
| "start": 257, |
| "end": 267, |
| "text": "(Brown 91)", |
| "ref_id": null |
| }, |
| { |
| "start": 270, |
| "end": 280, |
| "text": "(Gale 91a)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "SPECIFICATION OF STRUCTURAL MATCHING PROBLEM", |
| "sec_num": null |
| }, |
| { |
| "text": "A pair of Ja.panese and English sentences are parsed independently into (disjuuctive)feature structures. For our present purpose, a part of a feature structure is taken out as a dependency structure consisting of the content words 2 that appear in the original sentence. Ambiguity is represented by disjunctive feature structures (Kasper 87) . Since any relation other than modifier-modifyee dependencies is not considered here, path equivalence is not taken into consideration. Both of va.lue disjunction and general disjunction are allowed.", |
| "cite_spans": [ |
| { |
| "start": 330, |
| "end": 341, |
| "text": "(Kasper 87)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "We are currently using LFG-like grammars for both Japanese and English, where the value of the 'pred' label in an f-structure is the content word that is the head of the corresponding c-structure.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "We start with the definitions of simplified disjunctive feature structures, and then disjunctive dependency structures, that are extracted from the disjunctive feature structures obtained by the parsing process. 2. /Jr4 --41A 42, then labels in tl(41) and labels in tl(4_,) are sibling.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "3. If \u00a2 --41 V 42, then labels in 41 and labels in 42 are not sibling.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "Note that the sibling relation is not an equivalence relation. We refer to a set of feature labels in \u00a2 that are mutually sibling as a sibling label set of 4. Now, we are ready to define a dependency structure (DS).", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "Definition 4 A dependency structure ~b is an FS lhaI satisfies the following condition:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "Condition: Every sibling label set of \u00a2 includes exactly one 'pred' label.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "The idea behind those are that the value of a 'pred' label is a content word appearing in the original sentence, and that a sibling label set defines the dependency relation between content words. Among the labels in a sibling label set, the values of the labels other than 'pred' are dependent on (i.e., modify) the value of the 'pred' label. A DS can be drawn as a tree structure where the nodes are either a content word or disjunction operator and the edges represent the dependency relation. 3. If \u00a2 ----l : \u00a21, then sub(t1) are substructures of \u00a2.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "It\" ]f \u00a2 ----(~1 A (/)2, then for a~y (q C sub(el) and for any \u00a22 e sub(C2), \u00a21A\u00a2~ is a subslruclure oft.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "5. If \u00a2 = \u00a21 V \u00a22, then for for any '/r/)l ~ sub(~) 1 ) and for any \u00a22 E sub(C2), \u00a21 v\u00a22 is a sub-slr~ucture of \u00a2.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "The DS derived fi'om an FS is the maximuln substructure of the FS that satisfies the condition in Definition 4. The DS is uniquely determined fi'oln an FS.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "Definition 6 A disjunction-free maximal substructure of an FS \u00a2 is called a complete FS of \u00a2.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "An FS does not usually have a unique complete FS. This concept is important since the selection of a complete FS corresponds to alnbiguity resolution. Naturally, a lnaximal disjunction-free substructure of a DS \u00a2 is again a DS and is called a complete DS of \u00a2.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "DEFINITIONS OF DATA STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "Note that a substructure of a DS is not necessarily a DS. This is why the definition requires the condition in Definition 4.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Definition 7 A semi-complete DS of a DS \u00a2 is a substruclure of a complete DS of\u00a2 thai satisfies the condition in Definilion ~.", |
| "sec_num": null |
| }, |
| { |
| "text": "A complete DS ~/., can be decomposed into a set of non-overlapping selni-complete DSs. Such a decomposition defines the units of structural lnatching and plays the key role in our problem.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Definition 7 A semi-complete DS of a DS \u00a2 is a substruclure of a complete DS of\u00a2 thai satisfies the condition in Definilion ~.", |
| "sec_num": null |
| }, |
| { |
| "text": "Definition 8 A set of semi-complete DS of a DS \u00a2, D = {\u00a21,\"'\u00a2n}, is called a decomposition of \u00a2, iff every \u00a2i in the set contains at least one occurrence of 'pred' feature label, and every content word at the 'pred' feature label appeariT~g in '\u00a2 is contained in exactly one ~i.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Definition 7 A semi-complete DS of a DS \u00a2 is a substruclure of a complete DS of\u00a2 thai satisfies the condition in Definilion ~.", |
| "sec_num": null |
| }, |
| { |
| "text": "Definition 9 Th.e reduced DS of a DS (, with respect to a decomposition D = {\u00a21,\"-4',~} is constracted as follows: I. \u00a2i is transformed to a DS, \"pred : St', where Si is the set of all coT~le~l words appeari~J 9 i7~ \u00a2i. Th.is DS is referred to as red(it).", |
| "cite_spans": [ |
| { |
| "start": 141, |
| "end": 163, |
| "text": "DS, \"pred : St', where", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Definition 7 A semi-complete DS of a DS \u00a2 is a substruclure of a complete DS of\u00a2 thai satisfies the condition in Definilion ~.", |
| "sec_num": null |
| }, |
| { |
| "text": "2. If there is a direcl dependency relatiol~ between two conient words wl and w~ that are in \u00a2i and tj (i 7~ j), lh.en lhe dependency relation is allotted between \u00a2i and l/,j.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Definition 7 A semi-complete DS of a DS \u00a2 is a substruclure of a complete DS of\u00a2 thai satisfies the condition in Definilion ~.", |
| "sec_num": null |
| }, |
| { |
| "text": "Although this definition should be described precisely, we leave it with this more intuitive description. Examples of dependency structures and reduced dependency structures are found in Figures 1, 2 and 3 , where the decompositions are indicated by circles.", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 187, |
| "end": 206, |
| "text": "Figures 1, 2 and 3", |
| "ref_id": "FIGREF0" |
| } |
| ], |
| "eq_spans": [], |
| "section": "Definition 7 A semi-complete DS of a DS \u00a2 is a substruclure of a complete DS of\u00a2 thai satisfies the condition in Definilion ~.", |
| "sec_num": null |
| }, |
| { |
| "text": "It is not difficult to show that the reduced DS satisfies the condition of Definition 4.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Definition 7 A semi-complete DS of a DS \u00a2 is a substruclure of a complete DS of\u00a2 thai satisfies the condition in Definilion ~.", |
| "sec_num": null |
| }, |
| { |
| "text": "Structural matching problem of bilingual sentences is now defined formally.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "MATCHING OF BILIN-GUAL DEPENDENCY STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "Parsing parallel English and Japanese sentences results in feature structures, from which dependency structures are derived by removing unrelated features.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "MATCHING OF BILIN-GUAL DEPENDENCY STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "Assmne that ~.'E and 'OJ are dependency structures of English and Japanese sentences. The structural matching is to find the most plausible one-toone mapping between a decomposition of a complete DS of CE and a decomposition of a complete DS of C j, provided that the reduced DS of CE and the reduced DS of Cj w.r.t, the decompositions are isomorphic over the dependency relation. The isomorphism imposes a. natural one-to-one correspondence on the dependency relations between the reduced DSs.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "MATCHING OF BILIN-GUAL DEPENDENCY STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "Generally, the mapping need not always be oneto-one, i.e., all elements in a decomposition need not map into another decomposition. When the mapping is not one-to-one, we assume that dummy nodes are inserted in the dependency structures so that the mapping naturally extends to be one-toone.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "MATCHING OF BILIN-GUAL DEPENDENCY STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "When the decompositions of parallel sentences have such an isomorphic one-to-one mapping, we assume that there are systematic methods to compute similarity between corresponding elements in the decompositions and to compute similarity between the corresponding dependency relations 3.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "MATCHING OF BILIN-GUAL DEPENDENCY STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "We write the function defining the former similarity as f, and that of the latter as g. Then, f is a flmction over semi-complete DSs derived fi'om English and Japanese parallel sentences into a real number, and 9 is a function over feature label sets 3in the case of similarity between dependency relations, the original feature labels are taken into accotult.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "MATCHING OF BILIN-GUAL DEPENDENCY STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "of English and Japanese into a real number.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "MATCHING OF BILIN-GUAL DEPENDENCY STRUCTURES", |
| "sec_num": null |
| }, |
| { |
| "text": "and DS,,, of two languages, tile structural matching problem is to find an isomorphic oT~e-to-one mapping m be*ween decompositions of DSa aT~d DS2 that maximizes the sum of the vahtes of similarity functions, f and g.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Definition 10 Given dependency structures, DS1", |
| "sec_num": null |
| }, |
| { |
| "text": "That is, the problem is to find the fltnctioT~ m that maximizes ~-~m(f( d, re(d) The similarity functions can be defined in various ways. \"vVe assume some similarity measure between Japanese and English words. For instance, we assume that the similarity function f satisfies the following principles:", |
| "cite_spans": [ |
| { |
| "start": 64, |
| "end": 80, |
| "text": "~-~m(f( d, re(d)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Definition 10 Given dependency structures, DS1", |
| "sec_num": null |
| }, |
| { |
| "text": "1. f is a simple function defined by the similarity measure between content words of two la.nguages. 2. Fine-grained decompositions get larger similarity measure than coarse-grained decompositions. 3. Dummy nodes should give solne negative vahte to f.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Definition 10 Given dependency structures, DS1", |
| "sec_num": null |
| }, |
| { |
| "text": "The first principle is to simplify the complexity of the structural matching a.lgorithm. The second is to obtain detailed structural matching between parallel sentences and to avoid trivial results, e.g., the whole DSs are matched. The third is to avoid the introduction of dunnny nodes when it, is possible.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Definition 10 Given dependency structures, DS1", |
| "sec_num": null |
| }, |
| { |
| "text": "The fimction g should be defined according to the language pair. Although feature labels represent grammatical relation between content words or phrases and may provide useful information for measuring similarity, we do not use tile information at, our current stage. The reason is that we found it difficult to have a clear view on the relationship between feature labels of English and Japanese and on the meaning of feature labels between semi-complete dependency structures.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Definition 10 Given dependency structures, DS1", |
| "sec_num": null |
| }, |
| { |
| "text": "Tile structural matching of two dependency structures are combinatorially diflicult problem. V~re apply the 1)ranch-and-bound method to solve tile problem.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "STRUCTURAL MATCHING ALGORITHM", |
| "sec_num": null |
| }, |
| { |
| "text": "Tile branch-and-bound algorithm is a top-down depth-first backtracking algorithm for search problems. It looks for tile answers with the BEST score. Ill each new step, it estimates tile maximum value of the expected scores along the current path and compares it, with the currently known best score. The maxinmm expected score is usually calculated by a. simplified problem that guarantees to give a value not less than the best score attainable along the current path. If the maximuna expectation is less than the currently known best score, it means that there is no chance to find better answers by pursuing the path. Then, it gives up tile current path and hacktracks to try remaining paths.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "STRUCTURAL MATCHING ALGORITHM", |
| "sec_num": null |
| }, |
| { |
| "text": "We regard a dependency structure as a tree structure that inchtdes disjunction (OR nodes), and call a content word and a dependency relation as a node and an edge, respectively. Then a semi-complete dependency structure corresponds to a connected subgraph in the tree.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "STRUCTURAL MATCHING ALGORITHM", |
| "sec_num": null |
| }, |
| { |
| "text": "The matching of two dependency trees starts from the top nodes and the matching process goes along edges of the trees. During the matching process, three types of nondeterminisln arise:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "STRUCTURAL MATCHING ALGORITHM", |
| "sec_num": null |
| }, |
| { |
| "text": "1. Selection of top-most subgraphs in both of the trees (i.e., selection of a semi-complete DS) 2. Selection of edges ill both of tile trees to decide the correspondence of dependency relations 3. Selection of one of the disjuncts a.t an 'OR' node While tile matching is done top-down, the exact score of the matched subgraphs is calculated using the similarity function f.4 When the matching process proceeds to the selection of the second type, it selects an edge in each of the dependency trees. The maximum expected score of matching the subtrees under the selected edges are calculated from the sets of content words in the subtrees. Tile calculation method of the maximum expected score is defined ill solne relation with the similarity function f.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "STRUCTURAL MATCHING ALGORITHM", |
| "sec_num": null |
| }, |
| { |
| "text": "Suppose h is the function that gives the maximum expected score of two subgraphs. Also, suppose B and P be the currently known best score 4~,Ve do not take into account the similarity measure between dependency relations as stated in the preceding section. and the total score of the already matched subgraphs, respectively. If s and t are the subgraphs under the selected edges and s' and t' are the whole relnailfing subgraphs, the matching under s and t will be undertaken fi, rther only when the following inequation holds: h(s,t) + h(s',t') ", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 528, |
| "end": 545, |
| "text": "h(s,t) + h(s',t')", |
| "ref_id": null |
| } |
| ], |
| "eq_spans": [], |
| "section": "STRUCTURAL MATCHING ALGORITHM", |
| "sec_num": null |
| }, |
| { |
| "text": "P +", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "STRUCTURAL MATCHING ALGORITHM", |
| "sec_num": null |
| }, |
| { |
| "text": "Any selection of edges that does not satisfy this inequality cannot provide better matching than the currently known best ones.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "> B", |
| "sec_num": null |
| }, |
| { |
| "text": "All of the three types of nondeterminism are simply treated as the nondeterminism in the algorithm.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "> B", |
| "sec_num": null |
| }, |
| { |
| "text": "The syntactic ambiguities in the dependency structures are resolved sponta.lmously when the matching with the best score is obtained.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "> B", |
| "sec_num": null |
| }, |
| { |
| "text": "We have tested the structural matching algorithm with 82 pairs of sample sentences randomly selected froln a Japanese-English dictionary.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "EXPERIMENTS", |
| "sec_num": null |
| }, |
| { |
| "text": "We used a machine readable Japanese-English dictionary (Shimizu 79) and Roget's thesaurus (Roget 11) to measure the silnilarity of pairs of content words, which are used to define the fimctiou f.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "EXPERIMENTS", |
| "sec_num": null |
| }, |
| { |
| "text": "Given a pair of Japanese and English sentences, we take two methods to lneasure the similarity between Japanese and English content words appearing in the sentences.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Similarity of word pairs", |
| "sec_num": null |
| }, |
| { |
| "text": "For each Japanese content word wj apl)earing in the Japanese sentence, we can find a set of translatable English words fl'om the Japanese-Ellglish dietionary. When the Japanese word is a. polysemous word, we select an English word fi'om each polysemous entry. Let CE] be the set of such translatable English words of wj. Suppose CE is the set of contents words in the English sentence. The translatable pairs of w j, Tp(u u), is de.fined as follows:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Similarity of word pairs", |
| "sec_num": null |
| }, |
| { |
| "text": "Tp(wj) = {(wj,'wE) ['we E CE., n C.'L,}", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Similarity of word pairs", |
| "sec_num": null |
| }, |
| { |
| "text": "We use Roget's thesaurus to measure similarity of other word pairs. Roget's t.hesaurtls is regarded as a tree structure where words are a.llocated at the leaves of the tree: For each Japanese content word 'wj appearing in tim Japanese sentence, we can define the set of translatable English words of wa, CEj. From each English word in the set., the minimum distance to each of the English content words appearing in the English sentence is measured. 5 This minimum distance defines the similarity between pairs of Japanese and English words.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Similarity of word pairs", |
| "sec_num": null |
| }, |
| { |
| "text": "We decided to use this similarity only for estimating dissimilarity between Japanese and English word pairs. We set a predetermined threshold distance. If the minimal distance exceeds the threshold, the exceeded distance is counted as the negative similarity.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Similarity of word pairs", |
| "sec_num": null |
| }, |
| { |
| "text": "The similarity of two words Wl and w2 appearing in the given pair of sentences, sim((wl, w~) ), is defined as follows:", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 80, |
| "end": 92, |
| "text": "sim((wl, w~)", |
| "ref_id": null |
| } |
| ], |
| "eq_spans": [], |
| "section": "Similarity of word pairs", |
| "sec_num": null |
| }, |
| { |
| "text": ") = 6", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Similarity of word pairs", |
| "sec_num": null |
| }, |
| { |
| "text": "(wl, w2) E Tp(wl) or ('w2, 'wx) E Tp w2-I~ (,w~, w.) ~t Tp(w~) and (w2, w~) ft Tp(w.,) and the distance between wl and w., exceeds the threshold by k. 0 otherwise", |
| "cite_spans": [ |
| { |
| "start": 43, |
| "end": 52, |
| "text": "(,w~, w.)", |
| "ref_id": null |
| }, |
| { |
| "start": 56, |
| "end": 62, |
| "text": "Tp(w~)", |
| "ref_id": null |
| }, |
| { |
| "start": 79, |
| "end": 86, |
| "text": "Tp(w.,)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Similarity of word pairs", |
| "sec_num": null |
| }, |
| { |
| "text": "The similarity between corresponding semicomplete DSs is defined based on the similarity between the content words. Suppose that s and t are semi-colnplete DSs to be matched, and that Vs and Vt are the sets of content words in s and t. Let A be the less larger set of l~ and Vt and B be the other (I A I<l B I). For each injection p from A into B, the set of word pairs D derived from p can be defined as follows. Now, we define the similarity fimction f over Japaaese and English semi-colnplete DSs to give the naa.xinmm value to the following expression for all possible injections:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Similarity of semi-complete DSs", |
| "sec_num": null |
| }, |
| { |
| "text": "( = max/ \u00d7 O.951vd+IVd -~ J", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Similarity of semi-complete DSs", |
| "sec_num": null |
| }, |
| { |
| "text": "The summation gives the maximuna sum of the similarity of the content words in s and t. 0.95 is the penalty when the semi-complete DSs with more than one content words are used in the matching. Figures 1, 2 and 3 shows the results of the structural matching algorithm, in which the translatable pairs obtained fi'om the Japanese-English dictionary are shown by the equations. We used 82 pairs of Japanese and English sentences appearing in a Japanese-English dictionary. The results were checked and examined in detail by hand. Some of the sentences are not parsable because of the limited coverage of our current grammars. Although 59 pairs of them are parsable, 6 out of them do not include correct parse results.", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 194, |
| "end": 212, |
| "text": "Figures 1, 2 and 3", |
| "ref_id": "FIGREF0" |
| } |
| ], |
| "eq_spans": [], |
| "section": "Similarity of semi-complete DSs", |
| "sec_num": null |
| }, |
| { |
| "text": "The structural matchi,lg algorithm with the setting described above is applied to the 53 pairs. The cases where the correct, matchilig is not included in the best rated answers are 6 out of them. The remaining 47 pairs include the correct matching, of which 31 pairs result in the correct matching uniquely. Tal)le 1 sumnaarizes tile results.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Similarity of semi-complete DSs", |
| "sec_num": null |
| }, |
| { |
| "text": "Although the number of sentences used in tile experiments is small, the result, shows that about two third of the pairs give the unique matching, in which every syntactic ambiguity is resolved.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "EVALUATION AND DISCUSSION", |
| "sec_num": null |
| }, |
| { |
| "text": "The cases where no correct matching was obtained needs be examined. Some sentences contain an idiomatic expression that has coml)letely different syntactic structures fl'om the sentence structure of the other. Such an expression will 110 way be matched correctly except that the whole structures are matched intact. Other cases are caused by complex sentences that include an embedded sentence. When the verbs at the roots of the dependency trees are irrelevant, extraordinary matchings are produced. We intend not to use our method to match complex or compound sentences as a whole. ~,\u00a5e will rather use our method to find structural matching between simple sentences or verb phrases of two languages.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "EVALUATION AND DISCUSSION", |
| "sec_num": null |
| }, |
| { |
| "text": "Tile matching problmn of complex sentences are regarded as a different problem though the similar technique is usable. We think that the scores of matched phrases will help to identify tile corresponding phrases when we match complex sentences.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "EVALUATION AND DISCUSSION", |
| "sec_num": null |
| }, |
| { |
| "text": "Taking the sources of other errors into consideration, possible improvements are:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "EVALUATION AND DISCUSSION", |
| "sec_num": null |
| }, |
| { |
| "text": "1. Enhancement of English and Japanese grammars for wider coverage and lower error rate. 2. Introduction of more precise similarity measurement of content words.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "EVALUATION AND DISCUSSION", |
| "sec_num": null |
| }, |
| { |
| "text": "\u2022 Feature labels, for estimating matching plausibility of dependency relations \u2022 Part of speech, for measuring matching plausibility of content words \u2022 Other grammatical information: mood, voice, etc.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Utilization of grammatical information:", |
| "sec_num": "3." |
| }, |
| { |
| "text": "The first two iml)rovements are undoubtedly important. As for the similarity measurement of content words, completely different approaches such as statistical methods may be useful to get good translatable pairs (Brown 90) , (Gale 91) .", |
| "cite_spans": [ |
| { |
| "start": 212, |
| "end": 222, |
| "text": "(Brown 90)", |
| "ref_id": null |
| }, |
| { |
| "start": 225, |
| "end": 234, |
| "text": "(Gale 91)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Utilization of grammatical information:", |
| "sec_num": "3." |
| }, |
| { |
| "text": "Various grammatical information is kept in the feature descriptions produced in the parsing process. However, we should be very prudent in using it. Since English and Japanese are grammatically quite different, some grammatical rela.tion may not be preserved between them. In Figure 3 , solid arrows and circles show the correct matching. While 'benefit' matches with the structure consisting of ' ,~,,~ ' and ' ~_.~ ~ ', their dependent words 'trade' and ' H~:~' modify them as a verb modifier and as a noun modifier, the grammatical relation of which are quite different.", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 276, |
| "end": 284, |
| "text": "Figure 3", |
| "ref_id": "FIGREF0" |
| } |
| ], |
| "eq_spans": [], |
| "section": "Utilization of grammatical information:", |
| "sec_num": "3." |
| }, |
| { |
| "text": "This example highlights another interesting point. Dotted arrows and circles show another matching with the salne highest score. In this case, 'japan' is taken as a verb. This rather strange interpretation insists that 'japan' matches with ' H~ ' and ' .~ 6 '. Since 'japan' as a verb has little selnantic relation with ' []:~ ' as a country, discrimination of part-of-speech seems to be useful. On the other hand, the correspondence between 'benefit' and ' ~,~ ' is found in their noun entry in the dictionary. Since 'benefit' is used as a verb in the sentence, taking part-of-speech into consideration may jeopardize the correct matching, either. The fact that the verb and noun usages of 'benefit' bear common concept implies that more precise similarity measurement will solve this particular probleln. Since the interpretations of the sample English sentences are in different mood, imperative and declarative, the mood of a. sentence is also usefnl to remove irrelevant interpretations.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Utilization of grammatical information:", |
| "sec_num": "3." |
| }, |
| { |
| "text": "The structural matchillg problem of parallel texts is formally defined and our current implementation and experilnents are introduced. Although the research is at the preliminary stage and has a. very simple setting, the experiments have shown a. nulnber of interesting results. The method is easily enhanced by ilnproving the gramnm.rs and by incorporating more accurate similarity measurement. Number of other researches of building tra.nslation dictionaries and of deterlnining similarity relationship between words are useful to improve our method.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "CONCLUSIONS", |
| "sec_num": null |
| }, |
| { |
| "text": "To extract useful information fl'om bilingual corpora, structural matching is inevitable for language pairs like English and Japanese that have quite different linguistic structure. Incidentally, we have found that this dissimilarity plays an important role in resolving syntactic ambiguities since the sources of anlbiguities in English and Japanese sentences are in many cases do not coincide (Utsuro 92) . We are currently working on extracting verbal case frames of Japanese fi'om the results of structural matching of a aal)anese-l~nglish corpus (Utsuro 93). The salne teclmique is naturally a.pplicable to acquire verbal case fi'ames of English as well. Another application we are envisaging is to extract translation pattern from the results of structural matching. We plan to work on possible improvements discussed in the preceding section, and will make large scale experiments using translated newspal~er articles, based on the phrase matching stra.t.egy.", |
| "cite_spans": [ |
| { |
| "start": 395, |
| "end": 406, |
| "text": "(Utsuro 92)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "CONCLUSIONS", |
| "sec_num": null |
| }, |
| { |
| "text": "The dlstaame between words is tile length of tile shortest path in the thesatu'us tree.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| } |
| ], |
| "back_matter": [ |
| { |
| "text": "This work is partly supported by the (-;rants from Ministry of Education, \"Knowledge Science\" (#03245103).", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "ACKNOWLEDGMENTS", |
| "sec_num": null |
| } |
| ], |
| "bib_entries": { |
| "BIBREF0": { |
| "ref_id": "b0", |
| "title": "A Statistical Approach to Machine Translation, Computalional Linguistics, Vo1", |
| "authors": [ |
| { |
| "first": "P", |
| "middle": [ |
| "F" |
| ], |
| "last": "Brown", |
| "suffix": "" |
| } |
| ], |
| "year": 1990, |
| "venue": "", |
| "volume": "16", |
| "issue": "", |
| "pages": "79--85", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Brown, P.F., et al., A Statistical Approach to Ma- chine Translation, Computalional Linguistics, Vo1.16, No.2, pp.79-85, 1990.", |
| "links": null |
| }, |
| "BIBREF1": { |
| "ref_id": "b1", |
| "title": "Aligning Sentences ill Parallel Corpora, ACL-91", |
| "authors": [ |
| { |
| "first": "P", |
| "middle": [ |
| "F" |
| ], |
| "last": "Brown", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [ |
| "C" |
| ], |
| "last": "Lai", |
| "suffix": "" |
| }, |
| { |
| "first": "R", |
| "middle": [ |
| "L" |
| ], |
| "last": "Mercer", |
| "suffix": "" |
| } |
| ], |
| "year": 1991, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "169--176", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Brown, P.F., Lai, J.C. and Mercer, R.L., Align- ing Sentences ill Parallel Corpora, ACL-91, pp.169-176, 1991.", |
| "links": null |
| }, |
| "BIBREF2": { |
| "ref_id": "b2", |
| "title": "Two Languages are More Iuformative than One, ACL-91", |
| "authors": [ |
| { |
| "first": "I", |
| "middle": [], |
| "last": "Dagan", |
| "suffix": "" |
| }, |
| { |
| "first": "A", |
| "middle": [], |
| "last": "Itai", |
| "suffix": "" |
| }, |
| { |
| "first": "U", |
| "middle": [], |
| "last": "Schwall", |
| "suffix": "" |
| } |
| ], |
| "year": 1991, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "130--137", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Dagan, I., Itai, A. and Schwall, U., Two Lan- guages are More Iuformative than One, ACL- 91, pp.130-137, 1991a.", |
| "links": null |
| }, |
| "BIBREF3": { |
| "ref_id": "b3", |
| "title": "Identifying Word Correspondences in Parallel Texts, '91 DARPA Speech and Natural Language Workshop", |
| "authors": [ |
| { |
| "first": "", |
| "middle": [ |
| "W A" |
| ], |
| "last": "Gale", |
| "suffix": "" |
| }, |
| { |
| "first": "K", |
| "middle": [ |
| "W" |
| ], |
| "last": "Church", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [ |
| "W A" |
| ], |
| "last": "Gale", |
| "suffix": "" |
| }, |
| { |
| "first": "K", |
| "middle": [ |
| "W" |
| ], |
| "last": "Church", |
| "suffix": "" |
| } |
| ], |
| "year": 1991, |
| "venue": "ACL-91", |
| "volume": "", |
| "issue": "", |
| "pages": "152--157", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Gale. W.A. and Church, K.W., A Program for Aligning Sentences in Bilingual Corpora, ACL-91, pp.177-184, 1991b. Gale. W.A. and Church, K.W., Identifying Word Correspondences in Parallel Texts, '91 DARPA Speech and Natural Language Work- shop, pp.152-157, 1991.", |
| "links": null |
| }, |
| "BIBREF4": { |
| "ref_id": "b4", |
| "title": "Learning Translation Templates froln Bilingual Text, COLING-92", |
| "authors": [ |
| { |
| "first": "H", |
| "middle": [], |
| "last": "Kaji", |
| "suffix": "" |
| }, |
| { |
| "first": "Y", |
| "middle": [], |
| "last": "Kida", |
| "suffix": "" |
| }, |
| { |
| "first": "Y", |
| "middle": [], |
| "last": "Morimoto", |
| "suffix": "" |
| } |
| ], |
| "year": 1992, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "672--678", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Kaji, H., Kida, Y., and Morimoto, Y., Learning Translation Templates froln Bilingual Text, COLING-92, pp.672-678, 1992.", |
| "links": null |
| }, |
| "BIBREF5": { |
| "ref_id": "b5", |
| "title": "A Unification Method for Disjunctive Feature Descriptions", |
| "authors": [ |
| { |
| "first": "R", |
| "middle": [], |
| "last": "Kasper", |
| "suffix": "" |
| } |
| ], |
| "year": 1987, |
| "venue": "", |
| "volume": "87", |
| "issue": "", |
| "pages": "235--242", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Kasper, R., A Unification Method for Disjunc- tive Feature Descriptions, ACL-87, pp.235- 242, 1987.", |
| "links": null |
| }, |
| "BIBREF6": { |
| "ref_id": "b6", |
| "title": "The BICORD System: Combining Lexical Information from Bilingual Corpora. and Machine Readable Dictionaries", |
| "authors": [ |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Klavans", |
| "suffix": "" |
| }, |
| { |
| "first": "E", |
| "middle": [], |
| "last": "Tzoukermann", |
| "suffix": "" |
| } |
| ], |
| "year": 1990, |
| "venue": "COLING-90", |
| "volume": "", |
| "issue": "", |
| "pages": "174--179", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Klavans, J. and Tzoukermann, E., The BICORD System: Combining Lexical Information from Bilingual Corpora. and Machine Readable Dic- tionaries, COLING-90, pp.174-179, 1990.", |
| "links": null |
| }, |
| "BIBREF8": { |
| "ref_id": "b8", |
| "title": "Roget's Thesaurus", |
| "authors": [ |
| { |
| "first": "S", |
| "middle": [ |
| "R" |
| ], |
| "last": "Roget", |
| "suffix": "" |
| } |
| ], |
| "year": 1911, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Roget, S.R., Roget's Thesaurus, Crowell Co., 1911.", |
| "links": null |
| }, |
| "BIBREF9": { |
| "ref_id": "b9", |
| "title": "The Textual Knowledge Bank: Design, Construction, Applications", |
| "authors": [ |
| { |
| "first": "V", |
| "middle": [], |
| "last": "Sadler", |
| "suffix": "" |
| } |
| ], |
| "year": 1991, |
| "venue": "Proc. h~ler-national Workshop on Fundamental Research for the Future Generation of Natural Language Processing (FGNLP)", |
| "volume": "", |
| "issue": "", |
| "pages": "17--32", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Sadler, V., The Textual Knowledge Bank: De- sign, Construction, Applications, Proc. h~ler- national Workshop on Fundamental Research for the Future Generation of Natural Language Processing (FGNLP), pp.17-32, Kyoto, Japan, 1991.", |
| "links": null |
| }, |
| "BIBREF10": { |
| "ref_id": "b10", |
| "title": "Japanese-English Dictionary, Kodansha", |
| "authors": [ |
| { |
| "first": "M", |
| "middle": [], |
| "last": "Shimizu", |
| "suffix": "" |
| } |
| ], |
| "year": 1979, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Shimizu, M., et al. (ed.), Japanese-English Dictio- nary, Kodansha, 1979.", |
| "links": null |
| }, |
| "BIBREF11": { |
| "ref_id": "b11", |
| "title": "Lexical Knowledge Acquisition from Bilingual Corpora", |
| "authors": [ |
| { |
| "first": "T", |
| "middle": [], |
| "last": "Utsuro", |
| "suffix": "" |
| }, |
| { |
| "first": "Y", |
| "middle": [], |
| "last": "Matsumoto", |
| "suffix": "" |
| }, |
| { |
| "first": "M", |
| "middle": [], |
| "last": "Nagao", |
| "suffix": "" |
| } |
| ], |
| "year": 1992, |
| "venue": "COLING-92", |
| "volume": "", |
| "issue": "", |
| "pages": "581--587", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Utsuro, T., Matsumoto, Y., and Nagao, M., Lexi- cal Knowledge Acquisition from Bilingual Cor- pora., COLING-92, pp.581-587, 1992.", |
| "links": null |
| }, |
| "BIBREF12": { |
| "ref_id": "b12", |
| "title": "Verbal Case Frame Acquisition from Bilingual Corpora", |
| "authors": [ |
| { |
| "first": "T", |
| "middle": [], |
| "last": "Utsuro", |
| "suffix": "" |
| }, |
| { |
| "first": "Y", |
| "middle": [], |
| "last": "Matsumoto", |
| "suffix": "" |
| }, |
| { |
| "first": "M", |
| "middle": [], |
| "last": "Nagao", |
| "suffix": "" |
| } |
| ], |
| "year": 1993, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Utsuro, T., Matsumoto, Y., a.nd Nagao, M., Ver- bal Case Frame Acquisition from Bilingual Corpora, to appear IJCAI-93, 1993.", |
| "links": null |
| } |
| }, |
| "ref_entries": { |
| "FIGREF0": { |
| "text": "o\u00b0\u00b0\u00b0\u00b0\u00b0o.\u00b0..O~\u00b0 oo'\" ...................................... ...\u00b0\u00b0.... ........ Example of structural matching, No.3", |
| "type_str": "figure", |
| "uris": null, |
| "num": null |
| }, |
| "FIGREF1": { |
| "text": ") + ~t g(l, ,n.(/))) where d varies over semi-complete DS of DS1 and l varies over feature labels in D,-q. 1.", |
| "type_str": "figure", |
| "uris": null, |
| "num": null |
| }, |
| "TABREF0": { |
| "text": "", |
| "html": null, |
| "num": null, |
| "content": "<table><tr><td>NIL</td><td/></tr><tr><td>a</td><td>where a E A</td></tr><tr><td>1:4</td><td>where l E L, 4EFS</td></tr><tr><td colspan=\"2\">\u00a2 A ~b where 4,\u00a2 E FS</td></tr><tr><td colspan=\"2\">C V g, where \u00a2,\u00a2 E FS</td></tr><tr><td colspan=\"2\">To define (Disjunctive) Depen.dency Structures</td></tr><tr><td colspan=\"2\">as a special case of an FS, we first require the fol-</td></tr><tr><td colspan=\"2\">lowing definitions.</td></tr><tr><td colspan=\"2\">Definition 2 Top label set of an FS \u00a2, written as</td></tr><tr><td colspan=\"2\">tl(\u00a2), is defined:</td></tr><tr><td colspan=\"2\">tl(41) U ?~l(42).</td></tr><tr><td colspan=\"2\">Definition 3 A relation 'sibling' between feature</td></tr><tr><td colspan=\"2\">labels in 4 is defined:</td></tr><tr><td colspan=\"2\">1. If4 -= l : 41, then l and labels in 41 are not</td></tr><tr><td colspan=\"2\">sibling, and sibling relation holding in 41 also</td></tr><tr><td colspan=\"2\">holds in 4.</td></tr><tr><td>Definition 1 Simple feature structures (FS) (L is</td><td/></tr><tr><td>the sel of feature labels, and A is the set of atomic</td><td/></tr><tr><td>values) are defined recursively:</td><td/></tr><tr><td>1 Matching of compound sentences are done by cutting</td><td/></tr><tr><td>them up into simple sentence fragments.</td><td/></tr><tr><td>2In the present system, llOUllS, l)FOtK~utls, verbs, adjec-</td><td/></tr><tr><td>tives, mad adverbs are regarded as content, words.</td><td/></tr></table>", |
| "type_str": "table" |
| }, |
| "TABREF2": { |
| "text": "Results of experiment, s Parsing J al)anese and English sent.enccs", |
| "html": null, |
| "num": null, |
| "content": "<table><tr><td>Number of sentences</td><td>82</td><td/></tr><tr><td>Parse failure</td><td>23</td><td/></tr><tr><td>Parsable</td><td>59</td><td/></tr><tr><td>Correct parsability</td><td/><td/></tr><tr><td>Correctpa.rse</td><td colspan=\"2\">] 53 ] 89.8%(53/59)</td></tr><tr><td>Incorrect parse</td><td>6</td><td>10.2% (6/59)</td></tr><tr><td colspan=\"3\">The match with tile best score includes</td></tr><tr><td>Correct matching</td><td>47</td><td>89% (47/53)</td></tr><tr><td>no correct naatching</td><td>6</td><td>11% (6/53)</td></tr><tr><td colspan=\"2\">Single correct matching 34</td><td>64% (34/53)</td></tr><tr><td colspan=\"2\">Results of the experiments</td><td/></tr></table>", |
| "type_str": "table" |
| } |
| } |
| } |
| } |
