| { | |
| "paper_id": "M95-1001", | |
| "header": { | |
| "generated_with": "S2ORC 1.0.0", | |
| "date_generated": "2023-01-19T03:12:43.645970Z" | |
| }, | |
| "title": "DESIGN OF THE MUC-6 EVALUATIO N", | |
| "authors": [ | |
| { | |
| "first": "Ralph", | |
| "middle": [], | |
| "last": "Grishma", | |
| "suffix": "", | |
| "affiliation": {}, | |
| "email": "[email protected]" | |
| }, | |
| { | |
| "first": "Beth", | |
| "middle": [], | |
| "last": "Sundhei", | |
| "suffix": "", | |
| "affiliation": {}, | |
| "email": "[email protected]" | |
| } | |
| ], | |
| "year": "", | |
| "venue": null, | |
| "identifiers": {}, | |
| "abstract": "The sixth in a series of \"Message Understanding Conferences \" , which are designed to promote and evaluate research in information extraction, was held last fall. MUC-6 introduced several innovation s over prior MUCs, most notably in the range of different tasks for which evaluations were conducted. We describe the development of the \"message understanding\" task over the course of the prior MUCs, som e of the motivations for the new format, and the steps which led up to the formal evaluation .) THE MUC EVALUATION S Last fall we completed the sixth in a series of Message Understanding Conferences, which have been organized by NRAD, the RDT&E division of the Naval Command, Control and Ocean Surveillance Cente r (formerly NOSC, the Naval Ocean Systems Center) with the support of DARPA, the Defense Advance d Research Projects Agency. This paper looks briefly at the history of these Conferences and then examine s the considerations which led to the structure of MUC-6. 2 The Message Understanding Conferences were initiated by NOSC to assess and to foster research on th e automated analysis of military messages containing textual information. Although called \"conferences\" , th e distinguishing characteristic of the MUCs are not the conferences themselves, but the evaluations to whic h participants must submit in order to be permitted to attend the conference. For each MUC, participatin g groups have been given sample messages and instructions on the type of information to be extracted, an d have developed a system to process such messages. Then, shortly before the conference, participants are given a set of test messages to be run through their system (without making any changes to the system) ; the output of each participant's system is then evaluated against a manually-prepared answer key. The MUCs have helped to define a program of research and development. DARPA has a number o f information science and technology programs which are driven in large part by regular evaluations. The", | |
| "pdf_parse": { | |
| "paper_id": "M95-1001", | |
| "_pdf_hash": "", | |
| "abstract": [ | |
| { | |
| "text": "The sixth in a series of \"Message Understanding Conferences \" , which are designed to promote and evaluate research in information extraction, was held last fall. MUC-6 introduced several innovation s over prior MUCs, most notably in the range of different tasks for which evaluations were conducted. We describe the development of the \"message understanding\" task over the course of the prior MUCs, som e of the motivations for the new format, and the steps which led up to the formal evaluation .) THE MUC EVALUATION S Last fall we completed the sixth in a series of Message Understanding Conferences, which have been organized by NRAD, the RDT&E division of the Naval Command, Control and Ocean Surveillance Cente r (formerly NOSC, the Naval Ocean Systems Center) with the support of DARPA, the Defense Advance d Research Projects Agency. This paper looks briefly at the history of these Conferences and then examine s the considerations which led to the structure of MUC-6. 2 The Message Understanding Conferences were initiated by NOSC to assess and to foster research on th e automated analysis of military messages containing textual information. Although called \"conferences\" , th e distinguishing characteristic of the MUCs are not the conferences themselves, but the evaluations to whic h participants must submit in order to be permitted to attend the conference. For each MUC, participatin g groups have been given sample messages and instructions on the type of information to be extracted, an d have developed a system to process such messages. Then, shortly before the conference, participants are given a set of test messages to be run through their system (without making any changes to the system) ; the output of each participant's system is then evaluated against a manually-prepared answer key. The MUCs have helped to define a program of research and development. DARPA has a number o f information science and technology programs which are driven in large part by regular evaluations. The", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Abstract", | |
| "sec_num": null | |
| } | |
| ], | |
| "body_text": [ | |
| { | |
| "text": "MUCs are notable, however, in that they have substantially shaped the research program in informatio n extraction and brought it to its current state . 3", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "MUC-1 (1987) was basically exploratory ; each group designed its own format for recording the informatio n in the document, and there was no formal evaluation . By MUC-2 (1989) , the task had crystalized as one of template filling . One receives a description of a class of events to be identified in the text ; for each of thes e events one must fill a template with information about the event . The template has slots for informatio n about the event, such as the type of event, the agent, the time and place, the effect, etc . For MUC-2, the template had 10 slots . Both MUC-1 and MUC-2 involved sanitized forms of military messages about nava l sightings and engagements .", | |
| "cite_spans": [ | |
| { | |
| "start": 164, | |
| "end": 176, | |
| "text": "MUC-2 (1989)", | |
| "ref_id": null | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PRIOR MUG S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The second MUC also worked out the details of the primary evaluation measures, recall and precision . To present it in simplest terms, suppose the answer key has Nkey filled slots ; and that a system fills Ncorrec t slots correctly and Nincorrect incorrectly (with some other slots possibly left unfilled) . The n For MUC-3 (1991), the task shifted to reports of terrorist events in Central and South America, as reporte d in articles provided by the Foreign Broadcast Information Service, and the template became somewhat mor e complex (18 slots) . A sample MUC-3 message and template is shown in Figure 1 . This same task was use d for MUC-4 (1992) , with a further small increase in template complexity (24 slots) . For MUC-1 through 4 , all the text was in upper case . MUC-5 (1993) , which was conducted as part of the Tipster program,' represented a substantial furthe r jump in task complexity . Two tasks were involved, international joint ventures and electronic circuit fabrication, in two languages, English and Japanese . In place of a single template, the joint venture task employe d 11 object types with a total of 47 slots for the output -double the number of slots defined for MUC-4 -an d the task documentation also doubled in size to over 40 pages in length . A sample article and correspondin g template for the MUC-5 English joint venture task are shown in Figures 2 and 3 . The text shown is all upper case, but (for the first time) the test materials contained mixed-case text as well .", | |
| "cite_spans": [ | |
| { | |
| "start": 638, | |
| "end": 650, | |
| "text": "MUC-4 (1992)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 774, | |
| "end": 786, | |
| "text": "MUC-5 (1993)", | |
| "ref_id": null | |
| } | |
| ], | |
| "ref_spans": [ | |
| { | |
| "start": 598, | |
| "end": 606, | |
| "text": "Figure 1", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 1378, | |
| "end": 1393, | |
| "text": "Figures 2 and 3", | |
| "ref_id": "FIGREF3" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "PRIOR MUG S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "One innovation of MUC-5 was the use of a nested structure of objects . In earlier MUCs, each event had been represented as a single template -in effect, a single record in a data base, with a large numbe r of attributes . This format proved awkward when an event had several participants (e .g ., several victims of a terrorist attack) and one wanted to record a set of facts about each participant . This sort of information could be much more easily recorded in the hierarchical structure introduced for MUC-5, in which there wa s a single object for an event, which pointed to a list of objects, one for each participant in the event .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PRIOR MUG S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The sample template in Figure 3 illustrates several of the other features which added to the complexity of the MUC-5 task . The TIE_UP_RELATIONSHIP object points to the ACTIVITY object, which in tur n points to the INDUSTRY object, which describes what the joint venture actually did . Within the INDUSTRY object, the PRODUCT/SERVICE slot has to list not just the specific product or service of the joint venture , but also a two-digit code for this product or service, based on the top-level classification of the Standar d Industrial Classification . The TIE_UP_RELATIONSHIP also pointed to an OWNERSHIP object, which specified the total capitalization using standard codes for different currencies, and the percentage ownershi p of the various participants in the joint venture (which may involve some calculation, as in the example show n here) . While each individual feature of the template structure adds to the value of the extracted information , the net effect was a substantial investment by each participant in implementing the many details of the task . ", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 23, | |
| "end": 31, | |
| "text": "Figure 3", | |
| "ref_id": "FIGREF4" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "PRIOR MUG S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "DARPA convened a meeting of Tipster participants and government representatives in December 1993 t o define goals and tasks for MUC-6 . 5 Among the goals which were identified wer e", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "MUC-6 : INITIAL GOAL S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "\u2022 demonstrating domain-independent component technologies of information extraction which would b e immediately usefu l", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "MUC-6 : INITIAL GOAL S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "\u2022 encouraging work to make information extraction systems more portabl e", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "MUC-6 : INITIAL GOAL S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "\u2022 encouraging work on \" deeper understanding \"", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "MUC-6 : INITIAL GOAL S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Each of these can been seen in part as a reaction to the trends in the prior MUCs . The MUC-5 tasks, i n particular, had been quite complex and a great effort had been invested by the government in preparing th e training and test data and by the participants in adapting their systems for these tasks . Most participants worked on the tasks for 6 months ; a few (the Tipster contractors) had been at work on the tasks for considerably longer . While the performance of some systems was quite impressive (the best got 57% recall, 64 % precision overall, with 73% recall and 74% precision on the 4 \"core\" object types), the question naturally arose as to whether there were many applications for which an investment of one or several developers ove r half-a-year (or more) could be justified . Furthermore, while so much effort had been expended, a large portion was specific to the particular tasks . It wasn't clear whether much progress was being made on the underlying technologies which would be neede d for better understanding .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "MUC-6 : INITIAL GOAL S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "To address these goals, the meeting formulated an ambitious menu of tasks for MUC-6, with the idea that individual participants could choose a subset of these tasks . We consider the three goals in the three section s below, and describe the tasks which were developed to address each goal . ", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "MUC-6 : INITIAL GOAL S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The first goal was to identify, from the component technologies being developed for information extraction , functions which would be of practical use, would be largely domain independent, and could in the near ter m be performed automatically with high accuracy . To meet this goal the committee developed the \"name d entity\" task, which basically involves identifying the names of all the people, organizations, and geographi c locations in a text . The final task specification, which also involved time, currency, and percentage expressions, used SGM L markup to identify the names in a text . Figure 4 shows a sample sentence with named entity annotations . The tag ENAMEX (\"entity name expression\") is used for both people and organization names ; the tag NUME X (\"numeric expression\") is used for currency and percentages .", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 598, | |
| "end": 606, | |
| "text": "Figure 4", | |
| "ref_id": "FIGREF5" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "SHORT-TERM SUBTASKS", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The second goal was to focus on portability in the information extraction task -the ability to rapidly retarge t a system to extract information about a different class of events . The committee felt that it was important to demonstrate that useful extraction systems could be created in a few weeks . To meet this goal, we decide d that the information extraction task for MUC-6 would have to involve a relatively simple template, mor e like MUC-2 than MUC-5 ; this was dubbed \" mini-MUC\" . In keeping with the hierarchical object structure introduced in MUC-5, it was envisioned that the mini-MUC would have an event-level object pointing t o objects representing the participants in the event (people, organizations, products, etc .), mediated perhap s by a \"relational\" level object .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PORTABILITY", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "To further increase portability, a proposal was made to standardize the lowest-level objects (for people , organizations, etc .), since these basic classes are involved in a wide variety of actions . In this way, MU C participants could develop code for these low-level objects once, and then use them with many different type s of events . These low-level objects were named \"template elements\" .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PORTABILITY", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "As the specification finally developed, the template element for organizations had six slots, for the maxima l organization name, any aliases, the type, a descriptive noun phrase, the locale (most specific location), an d country . Slots are filled only if information is explicitly given in the text (or, in the case of the country, can be inferred from an explicit locale) . The tex t", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PORTABILITY", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "We are striving to have a strong renewed creative partnership with Coca-Cola,\" Mr . Dooner says . However, odds of that happening are slim since word from Coke headquarters in Atlanta is that . . . Ever on the lookout for additional evaluation measures, the committee decided to make the creation o f template elements for all the people and organizations in a text a separate MUC task . Like the named entity task, this was also seen as a potential demonstration of the ability of systems to perform a useful, relativel y domain independent task with near-term extraction technology (although it was recognized as being mor e difficult than named entity, since it required merging information from several places in the text) . The oldstyle MUC information extraction task, based on a description of a particular class of events (a \"scenario\" ) was called the \"scenario template\" task . A sample scenario template is shown in the appendix .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PORTABILITY", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Another concern which was noted about the MUCs is that the systems were tending towards relatively shallo w understanding techniques (based primarily on local pattern matching), and that not enough work was bein g done to build up the mechanisms needed for deeper understanding . Therefore, the committee, with strong encouragement from DARPA, included three MUC tasks which were intended to measure aspects of th e internal processing of an information extraction or language understanding system . These three tasks, which were collectively called SemEval ( \"Semantic Evaluation\") were :", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "MEASURES OF DEEP UNDERSTANDIN G", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "\u2022 Coreference : the system would have to mark coreferential noun phrases (the initial specificatio n envisioned marking set-subset, part-whole, and other relations, in addition to identity relations)", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "MEASURES OF DEEP UNDERSTANDIN G", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "\u2022 Word sense disambiguation : for each open class word (noun, verb, adjective, adverb) in the text , the system would have to determine its sense using the Wordnet classification (its \"synset\", in Wordne t terminology)", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "MEASURES OF DEEP UNDERSTANDIN G", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "\u2022 Predicate-argument structure : the system would have to create a tree interrelating the constituents of the sentence, using some set of grammatical functional relation s", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "MEASURES OF DEEP UNDERSTANDIN G", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The committee recognized that, in selecting such internal measures, it was making some presumptions regarding the structures and decisions which an analyzer should make in understanding a document . Not everyone would share these presumptions, but participants in the next MUC would be free to enter the informatio n extraction evaluation and skip some or all of these internal evaluations . Language understanding technolog y might develop in ways very different from those imagined by the committee, and these internal evaluation s might turn out to be irrelevant distractions . However, from the current perspective of most of the committee , these seemed fairly basic aspects of understanding, and so an experiment in evaluating them (and encouragin g improvement in them) would be worthwhile .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "MEASURES OF DEEP UNDERSTANDIN G", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The committee had proposed a very ambitious program of evaluations . We now had to reduce these proposal s to detailed specifications . The first step was to do some manual text annotation for the four tasksnamed entity and the SemEval triad -which were quite different from what had been tried before . Brief specifications were prepared for each task, and in the spring of 1994 a group of volunteers (mostly veterans of earlier MUCs) annotated a short newspaper article using each set of specifications . Problems arose with each of the SemEval tasks .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PREPARATION PROCES S Round 1 : Resolution of SemEval", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "\u2022 For coreference, there were problems identifying part-whole and set-subset relations, and distinguishin g the two (a proposal to tag more general coreference relations had been dropped earlier) ; a decision wa s later made to limit ourselves to identity relations .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PREPARATION PROCES S Round 1 : Resolution of SemEval", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "\u2022 For sense tagging, the annotators found that in some cases Wordnet made very fine distinctions an d that making these distinctions consistently in tagging was very difficult .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PREPARATION PROCES S Round 1 : Resolution of SemEval", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "\u2022 For predicate-argument structure, practically every new construct beyond simple clauses and nou n phrases raised new issues which had to be collectively resolved .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PREPARATION PROCES S Round 1 : Resolution of SemEval", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Beyond these individual problems, it was felt that the menu was simply too ambitious, and that w e would do better by concentrating on one element of the SemEval triad for MUC-6 ; at a meeting held in Maybe <COREF ID=\"136\" REF=\"134\">he</COREF>'ll even leave something from <COREF ID=\"138 \" REF=\"139\"><COREF ID=\"137\" REF=\"136\">his</COREF> office</COREF> for <COREF ID=\"140 \" REF=\"91\">Mr . Dooner</COREF> . Perhaps <COREF ID=\"144\">a framed page from the New Yor k Times, dated Dec . 8, 1987, showing a year-end chart of the stock market crash earlie r that year</COREF> . <COREF ID=\"141\" REF=\"137\">Mr . James</COREF> says <COREF ID=\"142 \" REF=\"141\">he</COREF> framed <COREF ID=\"143\" REF=\"144\" STATUS=\"OPT\">it</COREF> and kep t <COREF ID=\"145\" REF=\"144\">it</COREF> by <COREF ID=\"146\" REF=\"142\">his</COREF> desk as a \"personal reminder . It can all be gone like that . \" Figure 5 : Sample coreference annotation .", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 867, | |
| "end": 875, | |
| "text": "Figure 5", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "PREPARATION PROCES S Round 1 : Resolution of SemEval", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "June 1994, a decision was made to go with coreference . In part, this reflected a feeling that the problem s with the coreference specification were the most amenable to solution . It also reflected a conviction that coreference identification had been, and would remain, critical to success in information extraction, and so i t was important to encourage advances in coreference . In contrast, most extraction systems did not build ful l predicate-argument structures, and word-sense disambiguation played a relatively small role in extractio n (particularly since extraction systems operated in a narrow domain) .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PREPARATION PROCES S Round 1 : Resolution of SemEval", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The coreference task, like the named entity task, was annotated using SGML notation . A COREF tag has an ID attribute which identifies the tagged noun phrase or pronoun . It may also have an attribute of the form REF=n, which indicates that this phrase is coreferential with the phrase with ID n . Figure 5 shows an excerpt from an article, annotated for coreference . 6", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 298, | |
| "end": 306, | |
| "text": "Figure 5", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "PREPARATION PROCES S Round 1 : Resolution of SemEval", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Round 2 : annotatio n The next step was the preparation of a substantial training corpus for the two novel tasks which remaine d (named entity and coreference) . For annotation purposes, we wanted to use texts which could be redistribute d to other sites with minimal encumbrances . We therefore selected Wall Street Journal texts from 1987, 1988 , and 1989 which had already been distributed as part of the \"ACL/DCI \" CD-ROM and which were available at nominal cost from the Linguistic Data Consortium .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PREPARATION PROCES S Round 1 : Resolution of SemEval", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "SRA Corporation kindly provided tools which aided in the annotation process . Again a stalwart grou p of volunteer annotators was assembled ; 7 each was provided with 25 articles from the Wall Street Journal . There was some overlap between the articles assigned, so that we could measure the consistency of annotatio n between sites . This annotation was done in the winter of 1994-95 .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PREPARATION PROCES S Round 1 : Resolution of SemEval", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "A major role of the annotation process was to identify and resolve problems with the task specifications . For named entities, this was relatively straightforward . For coreference, it proved remarkably difficult to formulate guidelines which were reasonably precise and consistent . 8", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "PREPARATION PROCES S Round 1 : Resolution of SemEval", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Once the task specifications seemed reasonably stable, NRaD organized a \" dry run \" -a full-scale rehearsal for MUC-6, but with all results reported anonymously . The dry run took place in April 1995, with a scenario involving labor union contract negotiations, and texts which were again drawn from the 1987-89 Wall Stree t Journal . Of the sites which were involved in the annotation process, ten participated in the dry run . Results of the dry run were reported at the Tipster Phase II 12-month meeting in May 1995 .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Round 3 : dry run", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "An algorithm developed by the MITRE Corporation for MUC-6 was implemented by SAIC and used fo r scoring the coreference task [4] . The algorithm compares the equivalence classes defined by the coreferenc e links in the manually-generated answer key (the \"key\") and in the system-generated output (the \"response\") . The equivalence classes are the models of the identity equivalence coreference relation . Using a simpl e \u2022 Figure 6 : The participants in MUC-6 .", | |
| "cite_spans": [ | |
| { | |
| "start": 125, | |
| "end": 128, | |
| "text": "[4]", | |
| "ref_id": "BIBREF3" | |
| } | |
| ], | |
| "ref_spans": [ | |
| { | |
| "start": 423, | |
| "end": 431, | |
| "text": "Figure 6", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Round 3 : dry run", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "counting scheme, the algorithm obtains recall and precision scores by determining the minimal perturbations required to align the equivalence classes in the key and response .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Round 3 : dry run", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "A call for participation in the MUC-6 formal evaluation was issued in June 1995 ; the formal evaluation was held in September 1995 . The scenario definition was distributed at the beginning of September ; the test dat a was distributed four weeks later, with results due by the end of the week . The scenario involved changes i n corporate executive management personnel . The texts used for the formal evaluation were drawn from the 1993 and 1994 Wall Street Journal, an d were provided through the Linguistic Data Consortium . This data had been much less exposed than th e earlier Wall Street Journal data, and so was deemed suitable for the evaluation (participants were required t o promise not to look at Wall Street Journal data from this period during the evaluation) . There had originall y been consideration given to using a more varied test corpus, drawn from several news sources . It was decided , however, that multiple sources, with different formats and text mark-up, would be yet another complicatio n for the participants at a time when they were already dealing with multiple tasks .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "THE FORMAL EVALUATIO N", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "There were evaluations for four tasks : named entity, coreference, template element, and scenario template . There were 16 participants ; 15 participated in the named entity task, 7 in coreference, 11 in template element , and 9 in scenario template . The participants, and the tasks they participated in, are listed in Figure 6 .", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 320, | |
| "end": 328, | |
| "text": "Figure 6", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "THE FORMAL EVALUATIO N", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The results of the MUC-6 evaluations are described in detail in a companion paper in this volume , \"Overview of Results of the MUC-6 Evaluation\" . Overall, the evaluation met many, though not all, of th e goals which had been set by the initial planning conference in December of 1993 .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "THE FORMAL EVALUATIO N", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The named entity task exceeded our expectation in producing systems which could perform a relativel y simple task at levels good enough for immediate use . The nearly half the sites had recall and precision ove r 90% ; the highest-scoring system had a recall of 96% and a precision of 97% .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "THE FORMAL EVALUATIO N", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The template element task was harder and the scores correspondingly lower than for named entit y (ranging across most systems from 65 to 75% in recall, and from 75% to 85% in precision) . There seeme d general agreement, however, that having prepared code for template elements in advance did make it easie r to port a system to a new scenario in a few weeks . The goal for scenario templates -mini-MUC -wa s to demonstrate that effective information extraction systems could be created in a few weeks . Although it i s difficult to meaningfully compare results on different scenarios, the scores obtained by most systems after a few weeks (40% to 50% recall, 60% to 70% precision) were comparable to the best scores obtained in prio r MUCs .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "THE FORMAL EVALUATIO N", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Pushing improvements in the underlying technology was one of the goals of SemEval and its curren t survivor, coreference . Much of the energy for the current round, however, went into honing the definitio n of the task . We may hope that, once the task specification settles down, further evaluations, coupled with th e availability of coreference-annotated corpora, will encourage more work in this area .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "THE FORMAL EVALUATIO N", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Shown below is a sample filled template for the MUC-6 scenario template task . The scenario involved change s in corporate executive management personnel . For the tex t", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Appendix: Sample Scenario Templat e", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "McCann has initiated a new so-called global collaborative system, composed of world-wide accoun t directors paired with creative partners . In addition, Peter Kim was hired from WPP Group ' s J . Walter Thompson last September as vice chairman, chief strategy officer, world-wide .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Appendix: Sample Scenario Templat e", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "the following objects were to be generated : Although we cannot explain all the details of the template here, a few highlights should be noted . For eac h executive post, one generates a SUCCESSION_EVENT object, which contains references to the ORGANIZA -TION object for the organization involved, and the IN_AND_OUT object for the activity involving that pos t (if an article describes a person leaving and a person starting the same job, there will be two IN_AND_OU T objects) . The IN_AND_OUT object contains references to the objects for the PERSON and for the ORGANI-ZATION from which the person came (if he/she is starting a new job) . The PERSON and ORGANIZATIO N objects are the \"template element\" objects, which are invariant across scenarios .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Appendix: Sample Scenario Templat e", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "<SUCCESSION_EVENT", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Appendix: Sample Scenario Templat e", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Portions of this article are taken from the paper \"Message Understanding Conference-6 : A Brief History\", in COLING-96, Proc . of the Int'l Conj. on Computational Linguistics .2 The proceedings of earlier MUC conferences -for MUC-3, 4, and 5 -are available from Morgan Kaufmann Publishers , San Mateo, California .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The representatives of the research community were Jim Cowie, Ralph Grishman (committee chair), Jerry Hobbs, Pau l Jacobs, Len Schubert, Carl Weir, and Ralph Weischedel . The government people attending were George Doddington, Donna Harman, Boyan Onyshkevych, John Prange, Bill Schultheis, and Beth Sundheim .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The TYPE and MIN attributes which appear in the actual annotation have been omitted here for the sake of readability. 7 The annotation groups were from BBN, Brandeis Univ ., the Univ . of Durham, Lockheed-Martin, New Mexico State Univ . , NRaD, New York Univ ., PRC, the Univ . of Pennsylvania, SAIC (San Diego), SRA, SRI, the Univ . of Sheffield, Souther n Methodist Univ ., and Unisys .8 As experienced computational linguists, we probably should have known better than to think this was an easy task .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "", | |
| "sec_num": null | |
| } | |
| ], | |
| "back_matter": [], | |
| "bib_entries": { | |
| "BIBREF0": { | |
| "ref_id": "b0", | |
| "title": "General Semantic Patterns in Different Sublanguages", | |
| "authors": [ | |
| { | |
| "first": "E", | |
| "middle": [], | |
| "last": "Marsh", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1986, | |
| "venue": "Analyzing Language in Restricted Domains : Sublanguage Description and Processing", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Marsh, E . General Semantic Patterns in Different Sublanguages . In Analyzing Language in Restricted Do - mains : Sublanguage Description and Processing, R . Grishman and R . Kittredge, eds ., Lawrence Erlbau m Assoc ., Hillsdale, NJ, 1986 .", | |
| "links": null | |
| }, | |
| "BIBREF1": { | |
| "ref_id": "b1", | |
| "title": "Distinguishing Fact from Opinion and Events from Meta-Events", | |
| "authors": [ | |
| { | |
| "first": "C", | |
| "middle": [], | |
| "last": "Montgomery", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1983, | |
| "venue": "Proc . Conj.. Applie d Natural Language Processing", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Montgomery, C . Distinguishing Fact from Opinion and Events from Meta-Events . Proc . Conj.. Applie d Natural Language Processing, 1983 .", | |
| "links": null | |
| }, | |
| "BIBREF2": { | |
| "ref_id": "b2", | |
| "title": "Medical Language Processing : Computer Management of Narrative Data", | |
| "authors": [ | |
| { | |
| "first": "N", | |
| "middle": [], | |
| "last": "Sager", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "C", | |
| "middle": [], | |
| "last": "Friedman", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Lyman", | |
| "middle": [], | |
| "last": "", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "M", | |
| "middle": [], | |
| "last": "", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1987, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Sager, N ., Friedman, C ., and Lyman, M . et al . Medical Language Processing : Computer Management of Narrative Data . Addison-Wesley, Reading, MA, 1987 .", | |
| "links": null | |
| }, | |
| "BIBREF3": { | |
| "ref_id": "b3", | |
| "title": "A Model-Theoretic Coreference Scoring Scheme", | |
| "authors": [ | |
| { | |
| "first": "M", | |
| "middle": [], | |
| "last": "Vilain", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1996, | |
| "venue": "Proc . Sixth Message Understanding Conference (MUC-6)", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Vilain, M . et al ., A Model-Theoretic Coreference Scoring Scheme . Proc . Sixth Message Understanding Conference (MUC-6), Morgan Kaufmann, San Francisco, 1996 .", | |
| "links": null | |
| } | |
| }, | |
| "ref_entries": { | |
| "FIGREF1": { | |
| "type_str": "figure", | |
| "num": null, | |
| "uris": null, | |
| "text": "There were, however, a number of individual research efforts in information extraction underway before the first MUC , including the work on information formatting of medical narrative by Sager at New York University [3] ; the formatting of nava l equipment failure reports at the Naval Research Laboratory [1] ; and the DBG work by Montgomery for RADC (now Rome Labs) [2] .'Tipster is a U .S . Government program of research and development in the areas of information retrieval and informatio n extraction ." | |
| }, | |
| "FIGREF3": { | |
| "type_str": "figure", | |
| "num": null, | |
| "uris": null, | |
| "text": "A sample article from the MUC-5 English joint ventures task ." | |
| }, | |
| "FIGREF4": { | |
| "type_str": "figure", | |
| "num": null, | |
| "uris": null, | |
| "text": "A sample filled template from the MUC-5 English joint ventures task .Mr . <ENAMEX TYPE=\"PERSON\">Dooner</ENAMEX> met with <ENAMEX TYPE=\"PERSON\">Marti n Puris</ENAMEX>, president and chief executive officer of <ENAME X TYPE=\"ORGANIZATION\">Ammirati & Puris</ENAMEX>, about <ENAME X TYPE=\"ORGANIZATION\">McCann</ENAMEX>'s acquiring the agency with billings of <NUME X TYPE=\"MONEY\">$400 million</NUMEX>, but nothing has materialized ." | |
| }, | |
| "FIGREF5": { | |
| "type_str": "figure", | |
| "num": null, | |
| "uris": null, | |
| "text": "Sample named entity annotation ." | |
| }, | |
| "FIGREF6": { | |
| "type_str": "figure", | |
| "num": null, | |
| "uris": null, | |
| "text": "would yield an organization template element with five of the six slots filled : line identifies this as organization object 5 from article 9402240133) ." | |
| }, | |
| "TABREF0": { | |
| "num": null, | |
| "content": "<table><tr><td>TST1-MUC3-008 0</td></tr><tr><td>BOGOTA, 3 APR 90 (INRAVISION TELEVISION CADENA 1) -[REPORT] [JORGE ALONSO SIERR A</td></tr><tr><td>VALENCIA] [LAST WEEK, FEDERICO ESTRADA VELEZ HAD REJECTED TALKS BETWEEN THE GOVERN-</td></tr><tr><td>MENT AND THE DRUG TRAFFICKERS .</td></tr><tr><td>recall = N correct</td></tr><tr><td>precision =</td></tr></table>", | |
| "html": null, | |
| "text": "TEXT] LIBERAL SENATOR FEDERICO ESTRADA VELEZ WAS KIDNAPPED ON 3 APRIL AT THE CORNER OF 60TH AND 48TH STREETS IN WESTERN MEDELLIN, ONLY 100 ME-TERS FROM A METROPOLITAN POLICE CAI [IMMEDIATE ATTENTION CENTER] . THE ANTIO-QUTA DEPARTMENT LIBERAL PARTY LEADER HAD LEFT HIS HOUSE WITHOUT ANY BODY-GUARDS ONLY MINUTES EARLIER . AS HE WAITED FOR THE TRAFFIC LIGHT TO CHANGE , THREE HEAVILY ARMED MEN FORCED HIM TO GET OUT OF HIS CAR AND INTO A BLU E RENAULT . HOURS LATER, THROUGH ANONYMOUS TELEPHONE CALLS TO THE METROPOLITAN POLIC E AND TO THE MEDIA, THE EXTRADITABLES CLAIMED RESPONSIBILITY FOR THE KIDNAP-PING . IN THE CALLS, THEY ANNOUNCED THAT THEY WILL RELEASE THE SENATOR WITH A NEW MESSAGE FOR THE NATIONAL GOVERNMENT .", | |
| "type_str": "table" | |
| }, | |
| "TABREF1": { | |
| "num": null, | |
| "content": "<table><tr><td><DOCNO> 0592 </DOCNO ></td></tr><tr><td><DD> NOVEMBER 24, 1989, FRIDAY </DD ></td></tr><tr><td><SO> Copyright (c) 1989 Jiji Press Ltd . ; </SO ></td></tr><tr><td><TXT ></td></tr><tr><td>BRIDGESTONE SPORTS CO .</td></tr><tr><td>Figure 1 : A sample message and associated filled template from MUC-3 (terrorist domain) . Slots which are</td></tr><tr><td>not applicable to this type of incident (a kidnapping) are marked with an ' \" . For several of these slots ,</td></tr><tr><td>there are alternative \"correct\" answers ; only one of these answers is shown here .</td></tr></table>", | |
| "html": null, | |
| "text": "SAID FRIDAY IT HAS SET UP A JOINT VENTURE IN TAIWAN WIT H A LOCAL CONCERN ANDA JAPANESE TRADING HOUSE TO PRODUCE GOLF CLUBS TO B E SHIPPED TO JAPAN . THE JOINT VENTURE, BRIDGESTONE SPORTS TAIWAN CO ., CAPITALIZED AT 20 MILLIO N NEW TAIWAN DOLLARS, WILL START PRODUCTION IN JANUARY 1990 WITH PRODUCTION OF 20,000 IRON AND \"METAL WOOD\" CLUBS A MONTH . THE MONTHLY OUTPUT WILL B E LATER RAISED TO 50,000 UNITS, BRIDGESTON SPORTS OFFICIALS SAID . THE NEW COMPANY, BASED IN KAOHSIUNG, SOUTHERN TAIWAN, IS OWNED 75 PCT B Y BRIDGESTONE SPORTS, 15 PCT BY UNION PRECISION CASTING CO . OF TAIWAN AND TH E REMAINDER BY TAGA CO ., A COMPANY ACTIVE IN TRADING WITH TAIWAN, THE OFFICIAL S SAID .", | |
| "type_str": "table" | |
| } | |
| } | |
| } | |
| } |