| { | |
| "paper_id": "M92-1018", | |
| "header": { | |
| "generated_with": "S2ORC 1.0.0", | |
| "date_generated": "2023-01-19T03:13:06.199381Z" | |
| }, | |
| "title": "SRA SOLOMON : MUC-4 TEST RESULTS AND ANALYSI S", | |
| "authors": [ | |
| { | |
| "first": "Chinatsu", | |
| "middle": [], | |
| "last": "Aone", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Systems Research and Applications (SRA )", | |
| "institution": "", | |
| "location": { | |
| "addrLine": "2000 15th Street Nort h", | |
| "postCode": "2220", | |
| "settlement": "Arlington", | |
| "region": "VA" | |
| } | |
| }, | |
| "email": "[email protected]" | |
| }, | |
| { | |
| "first": "Doug", | |
| "middle": [], | |
| "last": "Mckee", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Systems Research and Applications (SRA )", | |
| "institution": "", | |
| "location": { | |
| "addrLine": "2000 15th Street Nort h", | |
| "postCode": "2220", | |
| "settlement": "Arlington", | |
| "region": "VA" | |
| } | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Sandy", | |
| "middle": [], | |
| "last": "Shinn", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Systems Research and Applications (SRA )", | |
| "institution": "", | |
| "location": { | |
| "addrLine": "2000 15th Street Nort h", | |
| "postCode": "2220", | |
| "settlement": "Arlington", | |
| "region": "VA" | |
| } | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Hatte", | |
| "middle": [], | |
| "last": "Bleje", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Systems Research and Applications (SRA )", | |
| "institution": "", | |
| "location": { | |
| "addrLine": "2000 15th Street Nort h", | |
| "postCode": "2220", | |
| "settlement": "Arlington", | |
| "region": "VA" | |
| } | |
| }, | |
| "email": "" | |
| } | |
| ], | |
| "year": "", | |
| "venue": null, | |
| "identifiers": {}, | |
| "abstract": "", | |
| "pdf_parse": { | |
| "paper_id": "M92-1018", | |
| "_pdf_hash": "", | |
| "abstract": [], | |
| "body_text": [ | |
| { | |
| "text": "In this paper, we report SRA's results on the MUC-4 task and describe how we trained our natural languag e processing system for MUC-4 . We also report on what worked, what didn't work, and lessons learned . Our MUC-4 system embeds the SOLOMON knowledge-based NLP shell which is designed for both domainindependence and language-independence. We are currently using SOLOMON for a Spanish and Japanes e text understanding project in a different domain . Although this was our first year participating in MUC, w e have built and are currently building other data extraction systems .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "INTRODUCTION", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Our TST3 and TST4 results are shown in Figures 1 and 2 . The similarity of these scores as well as thei r similarity to SRA-internal testing results reflects the portability of SRA's MUC-4 system . In fact, our scor e on the TST4 texts was better than that of TST3, even though those texts covered a different time perio d than that of the training texts or TST3 .", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 39, | |
| "end": 54, | |
| "text": "Figures 1 and 2", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "RESULTS", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Our matched-only precision and recall for both test sets were very high (TST3 : 68/47, TST4 : 73/49) . When SOLOMON recognized a MUC event, it did a very accurate and complete job at filling the requisit e templates .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "RESULTS", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "SOLOMON performance was tuned so that the all-templates recall and precision were as close as possibl e to maximize the F-Measure . As shown in Figure 3 , our F-Measure steadily increased over time . The fact that this slope has not yet leveled off shows SOLOMON's potential for improvement .", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 144, | |
| "end": 152, | |
| "text": "Figure 3", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "RESULTS", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "We spent a total of 9 staff months starting January 1, 1992 through May 31, 1992 on MUC-4 . A taskspecific breakdown of effort is shown in Figure 4 . The bulk of the work was spent porting SOLOMON t o a new domain with new vocabulary, concepts, template-output format, and fill rules . Approximately 72% of the effort was domain-dependent . However, about 63% of the total effort was language-independent, i .e . it would be directly applicable to understanding texts about terrorism in any language . We expect that our English MUC-4 system could be ported to a new language in about 3 months, given a basic grammar , lexicon and preprocessing data similar to the ones which existed for English . We partially demonstrated thi s ", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 139, | |
| "end": 147, | |
| "text": "Figure 4", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "EFFORT SPEN T", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "40% of the total effort building MUC-data was spent on lexicon and KB entry acquisition . Much of this dat a was acquired automatically. We used the supplied geographical data to automatically build location lexicon s and KBs . Using the development templates, we acquired lexical and KB entries for classes of domain term s such as human and physical targets and terrorist organizations . We automatically derived subcategorization information for the domain verbs from the development texts (cf. [1] ) . These automatically acquired lexicon s and KBs did require some manual cleanup and correction .", | |
| "cite_spans": [ | |
| { | |
| "start": 498, | |
| "end": 501, | |
| "text": "[1]", | |
| "ref_id": "BIBREF0" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Dat a", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Certain multi-word phenomena which occur frequently in texts but are unsuitable for general parsing wer e handled by pattern matching during Preprocessing . For example, we created patterns for Spanish phrases , complex location phrases, relative times, and names of political, military and terrorist organizations .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Dat a", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Modifications to SOLOMON's broad-coverage English grammar included adding more semantic restrictions, extending some phrase-structure rules, and improving general robustness .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Dat a", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Based on our knowledge engineering effort, we built a set of commonsense reasoning rules that are described in detail in our system description . Our EXTRACT module recognizes MUC-relevant events i n the output of SOLOMON and translates them into MUC-4 filled templates . We implemented all the domainspecific information as mapping rules or simple conversion functions (e .g . numeric values like \"at least 5 \" means \"5-\" ) . This data is stored in the knowledge base, and is completely language independent . ", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Dat a", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "We used TST2 texts for blind testing and the entire 1300 development texts for both testing and trainin g material . The development set was crucial to both our automated data acquisition and our knowledg e engineering task . We performed frequent testing to track and direct our progress . To raise recall, w e focussed on data acquisition ; to raise precision, we focussed on stricter definitions of \"legal\" MUC events .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "SYSTEM TRAININ G", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "To improve overall performance, we focussed on more robust syntactic and semantic analysis and mor e reliable event merging .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "SYSTEM TRAININ G", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The two main limiting factors were the number of development texts and templates and the amount of tim e allotted for the MUC-4 effort . With more texts, we could have applied other more data-intensive automate d acquisition techniques and had more examples of phenomena to draw upon . With more time, we would add more domain-dependent lexical knowledge and additional pragmatic inference rules . We also need to tune our EXTRACT mapping rules more finely and improve our discourse module for both NP reference an d event reference resolution . Integration of existing on-line resources such as machine-readable dictionaries , the World Factbook, or WordNet would also improve system performance. A more extensive testing and evaluation strategy at both the blackbox and glassbox levels would help direct progress, but was not feasibl e in the amount of time we had .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "LIMITING FACTOR S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "There were several areas where hybrid solutions worked very well . Totally automated knowledge acquisition was quite successful when supplemented by manual checking and editing of domain-crucial information . Similarly, augmenting a pure bottom-up parser with \"simulated top-down parsing\" (See SRA 's MUC-4 System Description) worked well . Improved Debris Semantics and significantly extended Pragmatic Inferencing wer e also important contributors to the system's performance .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "WHAT WAS OR WAS NOT SUCCESSFU L", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "SRA's SOLOMON NLP system has been designed for portability and proven to be highly reusable . Thi s includes portability to other domains, other languages, and other applications . As shown in Figure 5 , a larg e on-line resources are desirable . To ensure good results, it is necessary to have sufficient time for knowledg e engineering, testing and evaluation . Our experience underscores the fact that natural language understandin g is a highly data-driven problem . The system's performance is often proportional to the level of understandin g of the input and output . The MUC-4 development texts and templates were extremely helpful in this regard .", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 193, | |
| "end": 201, | |
| "text": "Figure 5", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "REUSABILITY", | |
| "sec_num": null | |
| } | |
| ], | |
| "back_matter": [ | |
| { | |
| "text": "Currently, our Spanish and Japanese data extraction project MURASAKI is using, without modification , the same processing modules and the core knowledge base as those used for MUC-4 . The MURASAKI system processes Spanish and Japanese language newspaper and journal articles as well as TV transcripts . This project's domain is the AIDS disease . Thus, the only difference between our MUC-4 system an d MURASAKI system is that the latter uses Spanish and Japanese lexicons, patterns and grammars, an d MURASAKI domain-dependent knowledge bases . SOLOMON has also been embedded in several Englis h message understanding systems : ALEXIS (operational) and WARBUCKS .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "annex", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "We have learned and reaffirmed the following points as the most crucial aspects of successful text understanding for data extraction .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "LESSONS LEARNED AND REAFFIRMED BY MUC-4", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "We must develop techniques and tools for acquiring timely, complete, and proven system data .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Overcoming the Knowledge Acquisition Bottleneck :", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "We need more robust, semantically constrained syntactic analysis . Grammars must be broad-coverage and highly accurate on complex input .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Solving the Parsing Problem :", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "We must handle real world discourse phenomena foun d in actual texts . The discourse architecture must be flexible enough to accommodate particular discours e phenomena which are crucial in particular domains or languages .MUC-4 has reaffirmed our knowledge of what is involved in porting an NLP system to a new domain . 9 staff months is a bare minimum for such an effort . Improved knowledge acquisition tools as well a s", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Developing Sophisticated Discourse Analysis :", | |
| "sec_num": null | |
| } | |
| ], | |
| "bib_entries": { | |
| "BIBREF0": { | |
| "ref_id": "b0", | |
| "title": "Using Statistics Gained from Corpora in a Knowledge-Based NL P System", | |
| "authors": [ | |
| { | |
| "first": "Doug", | |
| "middle": [], | |
| "last": "Mckee", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "John", | |
| "middle": [], | |
| "last": "Maloney", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1992, | |
| "venue": "Proceedings of The AAAI Workshop on Statistically-Based NLP Techniques", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Doug McKee and John Maloney. Using Statistics Gained from Corpora in a Knowledge-Based NL P System . In Proceedings of The AAAI Workshop on Statistically-Based NLP Techniques, 1992 .", | |
| "links": null | |
| } | |
| }, | |
| "ref_entries": { | |
| "TABREF1": { | |
| "content": "<table><tr><td>Processin g</td><td/><td/><td/><td/><td/><td/><td/><td/><td/><td/><td/><td/><td/><td/></tr><tr><td>I 0</td><td>I 100</td><td>I 200</td><td>I 300</td><td>I 400</td><td>I 500</td><td>I 600</td><td>I 700</td><td>I 500</td><td>I . 000</td><td>1000</td><td>1100</td><td>1 1200 1300</td><td>1 1400</td><td>moo</td></tr><tr><td>JAN 1</td><td/><td/><td/><td/><td/><td/><td colspan=\"3\">MAR 25 MAY 1</td><td/><td/><td>MAY17</td><td colspan=\"2\">MAY 31</td></tr><tr><td/><td/><td/><td/><td/><td/><td/><td colspan=\"3\">Hours of Effort</td><td/><td/><td/><td/><td/></tr><tr><td>Imo</td><td>11</td><td/><td>3125</td><td/><td>517</td><td/><td>5124</td><td/><td/><td>5125</td><td/><td>5127</td><td>5/3 1</td><td/></tr><tr><td>Noun</td><td>0</td><td/><td>300</td><td/><td>1240</td><td/><td>1380</td><td/><td/><td>1400</td><td/><td>1440</td><td>1500</td><td/></tr><tr><td>TST2</td><td>0</td><td/><td>11 .43</td><td/><td>19.48</td><td/><td>2625</td><td/><td/><td>27.43</td><td/><td>2525</td><td/><td/></tr><tr><td>TST3</td><td/><td/><td/><td/><td/><td/><td/><td/><td/><td/><td/><td/><td>2020</td><td/></tr><tr><td>T8T4</td><td/><td/><td/><td/><td/><td/><td/><td/><td/><td/><td/><td/><td>34 .14</td><td/></tr><tr><td/><td/><td/><td colspan=\"8\">Figure 3 : Tracking SOLOMON Performanc e</td><td/><td/><td/><td/></tr><tr><td/><td/><td/><td colspan=\"8\">Task Category~ % of Total Effort</td><td/><td/><td/><td/></tr><tr><td/><td/><td/><td>DATA</td><td/><td/><td/><td/><td/><td>7 1</td><td/><td/><td/><td/><td/></tr><tr><td/><td/><td/><td colspan=\"4\">Knowledge Engineering</td><td/><td/><td>1 3</td><td/><td/><td/><td/><td/></tr><tr><td/><td/><td/><td colspan=\"3\">Data Acquisition</td><td/><td/><td/><td>3 0</td><td/><td/><td/><td/><td/></tr><tr><td/><td/><td/><td colspan=\"2\">Grammar</td><td/><td/><td/><td/><td>7</td><td/><td/><td/><td/><td/></tr><tr><td/><td/><td/><td colspan=\"5\">Pragmatic Inference Rules</td><td/><td>1 1</td><td/><td/><td/><td/><td/></tr><tr><td/><td/><td/><td colspan=\"3\">Extract Data</td><td/><td/><td/><td>1 0</td><td/><td/><td/><td/><td/></tr><tr><td/><td/><td/><td colspan=\"3\">PROCESSING</td><td/><td/><td/><td>-2 9</td><td/><td/><td/><td/><td/></tr><tr><td/><td/><td/><td colspan=\"3\">Message Zoning</td><td/><td/><td/><td>3</td><td/><td/><td/><td/><td/></tr><tr><td/><td/><td/><td colspan=\"4\">Extract Extensions</td><td/><td/><td>7</td><td/><td/><td/><td/><td/></tr><tr><td/><td/><td/><td colspan=\"2\">Testing</td><td/><td/><td/><td/><td>1 0</td><td/><td/><td/><td/><td/></tr><tr><td/><td/><td/><td colspan=\"3\">Misc . Bug Fixing</td><td/><td/><td/><td>10</td><td/><td/><td/><td/><td/></tr><tr><td/><td/><td/><td colspan=\"8\">Figure 4 : Breakdown of Effort Spent for MUC-4</td><td/><td/><td/><td/></tr></table>", | |
| "html": null, | |
| "type_str": "table", | |
| "num": null, | |
| "text": "We spent 1 week porting our existing Message Zoner to deal with message headers in MUC messages . The Message Zoner could already recognize more general message structures such as paragraphs and sentences . We extended EXTRACT while maintaining domain and language independence of the module . Feature s added included event merging and handling of flat MUC templates instead of the more object-oriente d database records that SOLOMON is accustomed to . Our time spent on fixing bugs was distributed throughout the system, but problems in Debris Parsing and Debris Semantics received the most attention ." | |
| } | |
| } | |
| } | |
| } |