| { |
| "paper_id": "M95-1025", |
| "header": { |
| "generated_with": "S2ORC 1.0.0", |
| "date_generated": "2023-01-19T03:12:46.602400Z" |
| }, |
| "title": "", |
| "authors": [], |
| "year": "", |
| "venue": null, |
| "identifiers": {}, |
| "abstract": "", |
| "pdf_parse": { |
| "paper_id": "M95-1025", |
| "_pdf_hash": "", |
| "abstract": [], |
| "body_text": [ |
| { |
| "text": "Coreference Task Definitio n (V2.3, 8 Sep 95 ) 1 GENERAL NOTATION 1 .1 SGML Tagging 1 .2", |
| "cite_spans": [ |
| { |
| "start": 29, |
| "end": 35, |
| "text": "(V2.3,", |
| "ref_id": null |
| }, |
| { |
| "start": 36, |
| "end": 46, |
| "text": "8 Sep 95 )", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "The \"TYPE\" Attribute 1 .3", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "The \"ID\" and \"REF\" Attributes 1 .4", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "The \"MIN\" Attribute 1 .5", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "The \"STATUS\" Attribute 2 WHAT PART OF THE TEXT TO ANNOTATE 3 WHAT THINGS TO ANNOTATE 3 . Predicate Nominals and Time-dependent Identity 5 .5", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "Types and Tokens 5 .6", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "Functions and Values 5 .7 Metonymy 6 BASIS OF JUDGMENT 7 SCORING AND THE ORDERING OF LINKS 1 GENERAL NOTATIO N", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "The annotation for coreference is SGML tagging within the text stream . Referring expressions and their antecedents are tagged as follows :", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".SGML Taggin g", |
| "sec_num": "1" |
| }, |
| { |
| "text": "<COREF ID=\"100\">Lawson Mardon Group Ltd .</COREF> said <COREF ID=\"101 \" TYPE=\"IDENT\" REF=\"100\">it</COREF >", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".SGML Taggin g", |
| "sec_num": "1" |
| }, |
| { |
| "text": "The basic annotation contains the information to establish some type of link between an explicitly marked pair o f noun phrases . In the above example, the pronoun \"it\" is tagged as referring to the same entity as the phrase, \"Lawson Mardon Group Ltd . \"", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".SGML Taggin g", |
| "sec_num": "1" |
| }, |
| { |
| "text": "There is one markup per string . Other links can be inferred from the explicit links . We assume that the coreferenc e relation is symmetric and transitive, so if phrase A is marked as coreferential with B (indicated by a REF pointe r from A to B), we can infer that B is coreferential with A ; if A is coreferential with B, and B is coreferential with C , we can infer that A is coreferential with C .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".SGML Taggin g", |
| "sec_num": "1" |
| }, |
| { |
| "text": "The purpose of the TYPE attribute is to indicate the relationship between the anaphor and the antecedent. A t present only one such relationship, \"IDENT\" (for identity), is being annotated .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".2 The \"TYPE\" Attribute", |
| "sec_num": "1" |
| }, |
| { |
| "text": "The ID and REF attributes are used to indicate that there is a coreference link between two strings . The ID i s arbitrarily but uniquely assigned to the string during markup . The REF uses that ID to indicate the coreference link .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".3 The \"ID\" and \"REF\" Attribute s", |
| "sec_num": "1" |
| }, |
| { |
| "text": "The MIN attribute is used in the answer key (\"key\") to indicate the minimum string that the system unde r evaluation must include in the COREF tag in order to receive full credit for its output (\"response\") . So, in the nex t example, if the system response had omitted \"of Surrey, England\" from the COREF tag, the response woul d nonetheless receive full credit because it identified the minimum string . <COREF ID=\"100\" MIN=\"Haden MacLellan PLC\">Haden MacLellan PLC o f Surrey, England</COREF > <COREF ID=\"101\" TYPE=\"IDENT\" REF=\"100\">Haden MacLellan</COREF > Any response which includes the MIN string and does not include any tokens beyond those enclosed in th e <COREF> . . .</COREF> tags is valid . The MIN string will in general be the HEAD of the phrase ; see section 4 for a full discussion of this issue .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".4 The \"MIN\" Attribute", |
| "sec_num": "1" |
| }, |
| { |
| "text": "The STATUS (\"status\") attribute is used in the answer key when the markup is optional . The only value for this attribute is OPT (\"optional\") . The evaluation software will not score a string that is marked OPT in the key unles s the response has markup on that string . A potential example is given below. (It is marked OPT because a reader may not be certain that \"Livingston Street\" refers to the Board of Education .) Note that the optionality is marked only for the anaphor. <COREF ID=\"102\" MIN=\"Board of Education\">Our Board of Education</COREF > budget is just too high, the Mayor said . <COREF ID=\"103\" STATUS=\"OPT \" TYPE=\"IDENT\" REF=0102\">Livingston Street</COREF> has lost control .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".5 The \"STATUS\" Attribute", |
| "sec_num": "1" |
| }, |
| { |
| "text": "The <TXT> portion of the article should be annotated as well as the <HL>, the <DD>, and the <DATELINE> fro m the article header, but not any other lines from the header. (The DD tag sometimes doesn't appear at all, sometimes appears once, and sometimes appears twice . When it appears twice, only the SECOND instance is to be annotated . )", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "WHAT PART OF THE TEXT TO ANNOTAT E", |
| "sec_num": "2" |
| }, |
| { |
| "text": "Lines within the <TXT> portion of the article that start with the \"@\" sign signify a table or other special lin e formatting within the text and should NOT be annotated . (However, such lines may also appear within the <HL > portion of the article, and these should be annotated . )", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "WHAT PART OF THE TEXT TO ANNOTAT E", |
| "sec_num": "2" |
| }, |
| { |
| "text": "The coreference relation will be marked between elements of the following categories : NOUNS, NOUN PHRASES, and PRONOUNS . Elements of these categories are MARKABLES . PRONOUNS include both personal and demonstrative pronouns, and with respect to personal pronouns, all cases, including the possessive . Dates (\"January 23\"), currency expressions (\"$1 .2 billion\"), and percentages (\"17%\") are considered noun phrases .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".1 Markables", |
| "sec_num": "3" |
| }, |
| { |
| "text": "The relation is marked only between pairs of elements both of which are markables . This means that some markables that look anaphoric will not be coded, including pronouns, demonstratives, and definite NPs whos e antecedent is a clause rather than a markable . For example, i n Program trading is \"a racket,\" complains Edward Egnuss, a White Plains, N .Y ., investor and electronics sale s executive, \"and *it's not to the benefit of the small investor*, *that*'s for sure . \"", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".1 Markables", |
| "sec_num": "3" |
| }, |
| { |
| "text": "Though \"that\" is related to \"it's not to the benefit of the small investor\", the latter is not markable, so no anteceden t is annotated for \"that\" .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".1 Markables", |
| "sec_num": "3" |
| }, |
| { |
| "text": "Some indefinite NPs are not markables . See section 5 .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".1 Markables", |
| "sec_num": "3" |
| }, |
| { |
| "text": "Names and other Named Entities (as defined in the MUC-6 document titled \"Named Entity Task no relation is to be marked between \"1995\" and \"this year\" .", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 86, |
| "end": 90, |
| "text": "Task", |
| "ref_id": null |
| } |
| ], |
| "eq_spans": [], |
| "section": ".2 Names and Other Named Entitie s", |
| "sec_num": "3" |
| }, |
| { |
| "text": "Gerunds (verbal forms using a present participle) are not markable . I n *Slowing the economy* is supported by some Fed officials ; *it* is repudiated by others .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".Gerund s", |
| "sec_num": "3" |
| }, |
| { |
| "text": "one should not mark the relation between \"slowing the economy\" and \"it\" . A phrase headed by a present participl e is taken to be verbal if it can take an object (as in the above example) or can be modified by an adverb .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".Gerund s", |
| "sec_num": "3" |
| }, |
| { |
| "text": "Present participles which are modified by other nouns or adjectives (\"program trading\", \"excessive spending\"), ar e preceded by \"the\" or are followed by an \"of' phrase (\"the slowing of the economy\") are to be considered noun-lik e and ARE markable .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".Gerund s", |
| "sec_num": "3" |
| }, |
| { |
| "text": "The possessive forms of pronouns used as determiners are markable . Thus in its chairperso n there are are two potential markables for relations : \"its\" and the entire NP, \"its chairperson\" . Similarly, in \"th e man's arm\", there are two markables .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".4 Pronouns", |
| "sec_num": "3" |
| }, |
| { |
| "text": "First, second, and third-person pronouns are all markable, so i n \"There is no business reason for *my* departure\", *he* added .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".4 Pronouns", |
| "sec_num": "3" |
| }, |
| { |
| "text": "\"my\" and \"he\" should be marked as coreferential . Reflexive pronouns are markable, so i n *He* shot *himself* with *his* revolver .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".4 Pronouns", |
| "sec_num": "3" |
| }, |
| { |
| "text": "\"He\", \"himself', and \"his\" should all be marked coreferential .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".4 Pronouns", |
| "sec_num": "3" |
| }, |
| { |
| "text": "Prenominal occurrences of nouns, e .g ., in compound nouns, are markable . Thus i n The price of *aluminum* siding has steadily increased, as the market for *aluminum* reacts to the strike i n Chile .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".5 Bare Nouns", |
| "sec_num": "3" |
| }, |
| { |
| "text": "the relation between the two occurrences of \"aluminum\" should be marked . Note this presupposes that the two occurrences co-refer; they do, they both refer to the type of material .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".5 Bare Nouns", |
| "sec_num": "3" |
| }, |
| { |
| "text": "While nouns in prenominal positions are markable, the noun which appears at the head of a noun phrase is no t separately markable --it is markable only as part of the entire noun phrase . Thus in the passag e Linguists are a strange bunch . Some linguists even like spinach .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".5 Bare Nouns", |
| "sec_num": "3" |
| }, |
| { |
| "text": "it would not be correct to link the two instances of \"linguists\" .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".5 Bare Nouns", |
| "sec_num": "3" |
| }, |
| { |
| "text": "Assume that English has no zero pronouns ; in other words, the empty string is not markable . I n Bill called John and spoke with him for an hour .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".6 Implicit Pronoun s", |
| "sec_num": "3" |
| }, |
| { |
| "text": "there is no relation between the implicit subject of \"spoke\" and \"Bill\" .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".6 Implicit Pronoun s", |
| "sec_num": "3" |
| }, |
| { |
| "text": "Do not code relations between a relative pronoun and the head it attaches to or the gap that it fills .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".6 Implicit Pronoun s", |
| "sec_num": "3" |
| }, |
| { |
| "text": "Noun phrases which contain two or more heads (as defined in section 4 .1) are NOT markable . This restriction i s imposed so that each markable can be identified by a unique contiguous head substring . Thus no coreference is t o be marked fo r The boys and girls enjoy their breakfast .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conjoined Noun Phrases", |
| "sec_num": "3.7" |
| }, |
| { |
| "text": "The individual conjuncts are markable if they are separately coreferential with other phrases : <COREF ID=\"1\">Edna Fribble</COREF> and <COREF ID=\"2\">Sam Morton</COREF > addressed the meeting yesterday . <COREF ID=\"3\" REF=\"1\" TYPE=\"IDENT \" MIN=\"Fribble\">Ms . Fribble</COREF> discussed coreference, and <CORE F ID=\"4\" REF=\"2\" TYPE=\"IDENT\" MIN=\"Morton\">Mr . Morton</COREF> discusse d unnamed entities .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conjoined Noun Phrases", |
| "sec_num": "3.7" |
| }, |
| { |
| "text": "If the conjuncts share modifiers, the coreference is optional : <COREF ID=\"1\" MIN=\"Fribble\">Ms . Fribble</COREF> was <COREF ID=\"2\" REF=\"1 \" TYPE=\"IDENT\" STATUS=\"OPT\">president</COREF> and <CORE F TYPE=\"IDENT\" STATUS=\"OPT\" MIN=\"CEO\"> CEO of Amalgamated ID=\"3 \" Text REF=\"1\" Processing Inc .</COREF >", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conjoined Noun Phrases", |
| "sec_num": "3.7" |
| }, |
| { |
| "text": "The task is defined in order to allow maximal latitude for systems in identifying markables, and to decouple th e evaluation from that of accurately parsing noun phrases . Accordingly, the string generated by a system to identify a markable must include the head of the markable (as defined below) and may include any additional text up tp a maximal noun phrase (as defined below) .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "HOW MUCH OF THE MARKABLE TO ANNOTAT E", |
| "sec_num": "4" |
| }, |
| { |
| "text": "In preparing the key, the text element to be enclosed in SGML tags is the maximal noun phrase ; the head will b e designated by the MIN attribute .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "HOW MUCH OF THE MARKABLE TO ANNOTAT E", |
| "sec_num": "4" |
| }, |
| { |
| "text": "[We expect that in the future it may be possible, when separate noun phrase bracketings are available, to automatically generate the maximal NP markup from a markup using only heads . ]", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "HOW MUCH OF THE MARKABLE TO ANNOTAT E", |
| "sec_num": "4" |
| }, |
| { |
| "text": "For most noun phrases, the head will be the main noun, without its left and right modifiers . <COREF MIN=\"task\" . . .>the coreference taslt</COREF > <COREF MIN=\"contract\" . . .>the last contract</COREF> you will ever ge t <COREF MIN=\"quantity\" . . .>a large quantity of sugar</COREF > <COREF MIN=\"tons\" . . .>about 200,000 tons of sugar</COREF > If the head is a name, the entire name is marked . This includes suffixes such as \"Sr .\", \"III\", etc . on personal name s and \"Corp.\" on organization names ; it does not include personal titles or any modifiers . We follow in this regard the rules for marking personal and organization names for the Named Entity task . <COREF MIN=\"Frederick F . Fernwhistle Jr .\" . . .>the Honorable Frederick F . Fernwhistle Jr .</COREF > <COREF MIN=\"Ford Motor Co .\" . . .>Ford Motor Co . of Dearborn, Michigan<COREF > <COREF MIN=\"Georg Rath\" . . .>Herr Dr . Georg Rath</COREF > In the case of location designators consisting of multiple names, each name is considered a separate unit (as in th e Named Entity task) and the head is generally the first of these names, with the others treated as modifiers of the first name :", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".1 Head of a Phrase", |
| "sec_num": "4" |
| }, |
| { |
| "text": "<COREF MIN=\"Newark\" . . .>Newark, New Jersey</COREF>", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".1 Head of a Phrase", |
| "sec_num": "4" |
| }, |
| { |
| "text": "Dates, currency amounts, and percentages are also treated as atomic units, as in the Named Entity task : <COREF MIN=\"December 7, 1941\" . . .> December 7, 1941, a day which wil l live in infamy,</COREF > <COREF MIN=\"$1 .2 million\" . . .>$l .2 million in crisp bills</COREF > <COREF MIN=\"20%\">20% of the shares</COREF > In the case of \"headless\" constructions, the \"head\" --for coreference purposes --shall be the last token of the nou n phrase preceding any prepositional phrases, relative clauses, and other \"right modifiers\" :", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".1 Head of a Phrase", |
| "sec_num": "4" |
| }, |
| { |
| "text": "<COREF MIN=\"seven\" . . .>seven of the best</COREF > <COREF MIN=\"five\" . . .>the five who were left standing</COREF > <COREF MIN=--\"youngest\" . . .>the six youngest</COREF > If the maximal noun phrase is the same as the head, the MIN need not be marked .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".1 Head of a Phrase", |
| "sec_num": "4" |
| }, |
| { |
| "text": "The maximal noun phrase includes all text which may be considered a modifier of the noun phrase . This include s (among other modifiers) appositional phrases, non-restrictive relative clauses, and prepositional phrases which ma y be viewed as modifiers of the noun phrase or of a containing clause : *Mr . Holland * *the senior of the executives who will assume Holland's duties * *the rumor that the war had ended * *Fred Frosty, the ice cream king of Tyson's Corner, * *the Penn Central Co ., which used to run a railroad, * XYZ Inc . formed *a joint venture with Sony * Note that in the fourth and fifth cases the final comma may be viewed as part of the NP, and so is included in th e maximal NP ; in the last case, \"with Sony\" could equally well be taken to modify \"venture\" or \"formed\", and so i s included as part of the maximal NP around \"venture\" . Note also that in the \"Fred Frosty\" example, there is a coreference between the entire noun phrase and the appositional phrase, \"the ice cream king of Tyson's Corner\" ; see section 5 .3 for a discussion of this construct .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".2 Maximal Noun Phrase", |
| "sec_num": "4" |
| }, |
| { |
| "text": "In the case of a pair of conjoined noun phrases with shared complements or modifiers, the maximal noun phrase s will NOT include the conjunct . The maximal NP for the first conjunct will include all of the NP up to th e conjunction ; the maximal NP for the second conjunct will include all of the NP following the conjunction : <COREF ID=\"1\" MIN=\"Fribble\">Ms . Fribble</COREF> was <COREF ID=\"2\" REF=\"1 \" TYPE=\"IDENT\" STATUS=\"OPT\">president</COREF> and <CORE F TYPE=\"IDENT\" STATUS=\"OPT\" MIN=\"CEO\"> CEO of Amalgamated ID=\"3 \" Text REF=\"1\" Processing Inc .</COREF >", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".2 Maximal Noun Phrase", |
| "sec_num": "4" |
| }, |
| { |
| "text": "If the only difference between the head and the maximal noun phrase is the presence of an article --the word \"the\" , \"a\", or \"an\" at the beginning of the noun phrase --the MIN need not be explicitly marked . (The scoring progra m will automatically strip leading articles before comparing strings . )", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".3 Exceptions : Articles", |
| "sec_num": "4" |
| }, |
| { |
| "text": "The basic criterion for linking two markables is whether they are coreferential . whether they refer to the sam e object, set, activity, etc . It is not a requirement that one of the markables is \"semantically dependent\" on the other , or is an anaphoric phrase.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".1 Basic Coreference", |
| "sec_num": "5" |
| }, |
| { |
| "text": "We also make a coreference link between a \"bound anaphor\" and the noun phrase which binds it (even though on e may argue that such elements are not coreferential in the usual sense) . Thus we would link a quantified noun phrase and a pronoun dependent on that quantification : *Most computational linguists* prefer *their* own parsers .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".2 Bound Anaphors", |
| "sec_num": "5" |
| }, |
| { |
| "text": "Note that a quantified noun phrase would also be linked to subsequent anaphors, outside the scope of quantification, through the usual relation of identity of coreference . Thug in the following text all three noun phrases would be linked: *Every TV network* reported *its* profits yesterday . *They* plan to release full quarterly statement s tomorrow .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".2 Bound Anaphors", |
| "sec_num": "5" |
| }, |
| { |
| "text": "By this rule, a pronoun in a relative clause which is bound to the head of the clause would get a coreference link t o the entire NP. Thus, fo r every man who knows his own min d we would establish a coreference link between \"his\" and the entire noun phrase \"every man who knows his ow n mind\" : <COREF ID=\"l\" MIN=\"man\">every man who knows <COREF ID=\"2\" REF=\"1 \" TYPE=\"IDENT\">his <COREF>own mind</COREF >", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".2 Bound Anaphors", |
| "sec_num": "5" |
| }, |
| { |
| "text": "A typical use of an appositional phrase is to provide an alternative description or name for an object : Julius Caesar, the well-known emperor ,", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".3 Appositio n", |
| "sec_num": "5" |
| }, |
| { |
| "text": "This identity of reference is to be represented by a coreference link between the appositional phrase, \"the wellknown emperor\" and the ENTIRE noun phrase, \"Julius Caesar, the well-known emperor\" : <COREF ID=\"l\" MIN=\"Julius Caesar\">Julius Caesar, <COREF ID=\"2\" REF=\"l \" MIN=\"emperor\" TYPE=\"IDENT\"> the well-known emperor,</COREF></COREF > The appositional phrase may be separated from the head by other modifiers . Thus Peter Holland, 45, deputy general manager, . . . becomes <COREF ID=\"l\" MIN=\"Peter Holland\">Peter Holland, 45, <COREF ID=\"2 \" REF=\"1\" TYPE=\"IDENT\" MIN=\"manager\"> deputy general manager,</COREF>< / COREF > Appositional phrases that are marked indefinite are NOT considered to be coreferential . Examples o f noncoreferential appositional phrases include the following : Ms . Ima Head, a 10-year MUC veteran , San Diego, one of America's finest cities, Currently, only appositional phrases that are overtly marked via punctuation are considered markables . Thus, n o coreference is marked in cases such as the following : *the real estate company* * Century 21 * *the realtor* *Century 21 * *presidential advisor* *Joe Smarty * *Treasury Secretary* *Bucks* *the job of *manager* *", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".3 Appositio n", |
| "sec_num": "5" |
| }, |
| { |
| "text": "Predicate nominals are also typically coreferential with the subject . Thus in the example Bill Clinton is the President of the United States .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".4 Predicate Nominals and Time-dependent Identit y", |
| "sec_num": "5" |
| }, |
| { |
| "text": "we would record a coreference link between \"Bill Clinton\" and \"the President of the United States\" . Coreference should NOT be recorded if the text only asserts the possibility of identity between two markables . I n Phinneas Flounder may be the dumbest man who ever lived . Phinneas Flounder is a leading candidate to become president . If elected, Phinneas Flounder would be the first Californian in the Oval Office .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".4 Predicate Nominals and Time-dependent Identit y", |
| "sec_num": "5" |
| }, |
| { |
| "text": "no coreference is to be recorded .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".4 Predicate Nominals and Time-dependent Identit y", |
| "sec_num": "5" |
| }, |
| { |
| "text": "Neither should coreference be recorded when the predicate nominative is marked indefinite . Examples of noncoreferential predicate nominatives includ e Mediation is a viable alternative to bankruptcy . Farm-debt mediation is one of the Farm Belt's success stories . ARPA program managers are nice people .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".4 Predicate Nominals and Time-dependent Identit y", |
| "sec_num": "5" |
| }, |
| { |
| "text": "Two markables should be recorded as coreferential if the text asserts them to be coreferential at ANY TIME . Thu s Henry Higgins, who was formerly sales director for Sudsy Soaps, became president of Dreamy Detergent s should be annotated a s <COREF ID=\"l\" MIN=\"Henry Higgins\">Henry Higgins, who was formerly <CORE F ID=\"2\" MIN=\"director\" REF=\"l\" TYPE=\"IDENT\">sales director for Suds y Soaps,</COREF></COREF> became <COREF ID=\"3\" MIN=\"president\" REF=\"1 \" TYPE=\"IDENT\">president of Dreamy Detergents</COREF > Even if the copula or inchoative verb is embedded, coreference should be marked, as i n Dreamy Detergents named Henry Higgins to be presiden t which should be annotated as Dreamy Detergents named <COREF ID=\"1\">Henry Higgins</COREF> to be <CORE F ID=\"2\" REF,---\"1\" TYPE=\"IDENT\">president</COREF > When the copula is implied by the semantics of the verb but is not expressed overtly, the coreference relation wil l be marked optional in the answer key . Expressions of equivalence involving the word \"as\" will also be marke d optional . The NPs enclosed in asterisks in the following examples will be marked optionally coreferential : Dreamy Detergents named *Henry Higgins* *president * *Henry Higgins* is considered *Sudsy Soap's best sales director* *Higgins* will serve as *president of Dreamy Detergents *", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".4 Predicate Nominals and Time-dependent Identit y", |
| "sec_num": "5" |
| }, |
| { |
| "text": "The general principle for annotating coreference is that two markables are coreferential if they both refer to sets , and the sets are identical, or they both refer to types, and the types are identical . There are a number of problemati c cases where one can argue whether something is a set or a type . There is no simple algorithm for determining th e ontological category of a referent . There are, though, some useful rules . Most occurrences of bare plurals refer to types or kinds, not to sets . In . . .*producers* don't like to see a hit wine increase in price . . . *Producers* have seen this market opening u p and *they*'re now creating wines that appeal to these people .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".Types and Token s", |
| "sec_num": "5" |
| }, |
| { |
| "text": "\"producers\", \"Producers\", and \"they\" refer to types and they all refer to the same type . Notice that if interpreted as referring to sets, they would not all refer to the same set . More properly, there is no reason to think they woul d corefer ; not all the producers who have seen the market opening up have created new wines .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".Types and Token s", |
| "sec_num": "5" |
| }, |
| { |
| "text": "Note that a type can be referred to by a bare plural, a definite singular np (\"the tiger is fast becoming extinct\") or a (bare) prenominal . I n The action followed by one day an Intelogic announcement that it will retain an investment banker to explor e alternatives \"to maximize *shareholder* value,\" including the possible sale of the company . Mr . Edelman declined to specify what prompted the recent moves, saying they are meant only to benefit *shareholders * when \"the company is on a roll . \" the two starred occurrences corefer to the type : shareholder (of Intelogic) .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".Types and Token s", |
| "sec_num": "5" |
| }, |
| { |
| "text": "In GM announced *its third quarter profit* . *It* was *$0 .02* .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".6 Functions and Value s", |
| "sec_num": "5" |
| }, |
| { |
| "text": "all three starred phrases refer to an amount of money ; they all refer to the same amount of money . Hence they are coreferential . The first phrase, in context, refers to that amount via referring to a function, say of companies an d quarters of a year--or times . (In addition, the \"its\" in the first NP would be linked to GM .) I n General Motors announced {their third quarter profit of *$0 .02* } .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".6 Functions and Value s", |
| "sec_num": "5" |
| }, |
| { |
| "text": "the bracketed and starred phrases are coreferential . They refer to one and the same amount of money . Note that here, as in the case of apposition, the result is that a phrase is marked as being coreferential with a part of the phrase. In *The temperature* is *90* . . . .The temperature is rising .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".6 Functions and Value s", |
| "sec_num": "5" |
| }, |
| { |
| "text": "the first occurrence of \"the temperature\" refers to the value of the function at arguments (places, times) supplied b y context . That occurrence is coreferential with \"90\" . In the second occurrence, \"the temperature\" refers to th e function (indirectly, by way of referring to the derivative of the function) . So it is not coreferential with the first occurrence or with \"90\" .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".6 Functions and Value s", |
| "sec_num": "5" |
| }, |
| { |
| "text": "There will be cases where a phrase could arguably refer to either a set or a type ; in such ambiguous cases, the coreference should be recorded but marked as optional .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".6 Functions and Value s", |
| "sec_num": "5" |
| }, |
| { |
| "text": "The pervasive phenomenon of metonymy raises a problem for Coreference relations . Do we annotate and recognize the relation before or after coercion? Here are some texts to consider :", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".7 Metonymy", |
| "sec_num": "5" |
| }, |
| { |
| "text": "(1) *The White House* sent its health care proposal to Congress yesterday . Senator Dole said *the administration*'s bill had little chance of passing .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".7 Metonymy", |
| "sec_num": "5" |
| }, |
| { |
| "text": "(2) *Ford* announced a new product line yesterday . *Ford* spokesman John Smith said *they* will start manufacturing widgets .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".7 Metonymy", |
| "sec_num": "5" |
| }, |
| { |
| "text": "(3) I bought the New York Times this morning . I read that the editor of the New York Times is resigning .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".7 Metonymy", |
| "sec_num": "5" |
| }, |
| { |
| "text": "(4) *The United States* is a democracy . *The United States* has an area of 3 .5 million square miles .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".7 Metonymy", |
| "sec_num": "5" |
| }, |
| { |
| "text": "We propose that coreference be determined with respect to coerced entities . Of course, this still leaves open the question as to the circumstances under which coercion is required . In (1) there is a coercion from the White House to the administration operating out of the White House, and that is IDENT with \"the administration\" ; so \"White House\" and \"administration\" are IDENT. (Notice that there is also a question as to whether the adminstration' s proposal is the same as its bill . This too requires a coercion of sorts .) In (2), while there might seem to be a coercion from Ford to a spokesman for Ford, we believe that such a coercion is not necessary, for it is plausible that corporations, as legal persons, can do many of the things that people can do--such as 'announce' . They may have to do some or all such things through other agents, but many people do many things that way . And if Ford ca n announce, then it, through one of its spokesmen, can \"say\" . Believing that no coercion is required, we would mar k as coreferential the first instance of \"Ford\", the second instance of \"Ford\" (in the phrase \"Ford spokesman Joh n Smith\"), and \"they\", but would NOT mark the phrase \"Ford spokesman John Smith\" as coreferential with anythin g else in this passage . In (3) the first \"New York Times\" is coerced into a copy of the paper published by the New York Times and the second is coerced into the organization ; so they are not IDENT. (4) is somewhat akin to (2) . Countries are both geographical entities and governmental units . Thus, no coercion is necessary and the two starre d occurrences are coreferential .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".7 Metonymy", |
| "sec_num": "5" |
| }, |
| { |
| "text": "In the absence of general principles, a body of such decisions will need to be developed to codify the rules fo r coercion and coreference . In cases where there has been no clear precedent, the answer keys for formal evaluation s will need to mark coreference as optional .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": ".7 Metonymy", |
| "sec_num": "5" |
| }, |
| { |
| "text": "The coreference judgments should be based on the intelligent reader's knowledge of the world resulting from his o r her best understanding of the text . It should not be based on a theory of the structure of the text, or on a linguisti c", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "BASIS OF JUDGMEN T", |
| "sec_num": "6" |
| } |
| ], |
| "back_matter": [ |
| { |
| "text": "theory of how NPs are resolved, or on estimates of what the typical NLP system could do . This means that some relations will be impossible for current NLP systems to recover, but this is why the task will push the technology . The annotators should assume that they are typical intelligent readers .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Appendix D : Coreference Task Definition (v2 .3)", |
| "sec_num": null |
| }, |
| { |
| "text": "If three markables, A, B, and C, are coreferential, this relationship could be recorded in the key in several ways : for example, by a REF pointer in both B and C pointing to A, or by a REF pointer in B pointing to A and a RE F pointer in C pointing to B . A similar range of variations is possible in a system response . The current scoring rule s provide that any correct key, when compared to any correct response, will yield a 100% recall/100% precisio n score, independent of the way the coreference relation is encoded in the key by REF pointers . However, if the response is incomplete, its recall score CAN be affected by the way in which the coreference relation is encoded b y the key. It is therefore recommended that each markable which participates in a coreference relation have a RE F pointer to the most recent prior coreferential markable which does not have STATUS=\"OPT\" .", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "SCORING AND THE ORDERING OF LINKS", |
| "sec_num": "7" |
| } |
| ], |
| "bib_entries": {}, |
| "ref_entries": {} |
| } |
| } |
