| { | |
| "paper_id": "M91-1008", | |
| "header": { | |
| "generated_with": "S2ORC 1.0.0", | |
| "date_generated": "2023-01-19T03:15:30.695620Z" | |
| }, | |
| "title": "GTE'S TEXT INTERPRETATION AID (TIA) : MUC-3 TEST RESULTS AND ANALYSI S", | |
| "authors": [ | |
| { | |
| "first": "Charles", | |
| "middle": [ | |
| "T" | |
| ], | |
| "last": "Taylor", | |
| "suffix": "", | |
| "affiliation": {}, | |
| "email": "" | |
| } | |
| ], | |
| "year": "", | |
| "venue": null, | |
| "identifiers": {}, | |
| "abstract": "GTE has actively participated in the Third Annual Message Understanding Conference (MUC-3) sponsored b y the Naval Ocean System Center (NOSC) for the past 7 months using its natural language processing system, TI A (Text Interpretation Aid)[5]. During this period, TIA system development consisted of two tasks: (1) lexicon tool development, and (2) tailoring TIA to meet the specific needs of the MUC-3 domain, i .e., terrorism in Lati n American countries. Lexicon tool development facilitates system (lexicon) adaptation to new domains (terrorism , drug interdiction, third world launchings, etc .) by semi-automating the manual task of entering words/phrases into TIA's lexicons. Tailoring the existing TIA system to parse and extract MUC-3 domain messages, has allowed GT E to participate in the MUC-3 conference tasks. The purpose of this paper is to report our techniques and findings during these 7 months of MUC-TIA syste m development. Additionally, GTE's official scores for Phase I and Phase II of MUC-3 will be disclosed and discussed. The remainder of this paper is divided into seven sections describing scoring results, justification and analysis o f scores, system development effort, limiting factors, training, reusability of the resultant MUC-TIA system, and fina l conclusions .", | |
| "pdf_parse": { | |
| "paper_id": "M91-1008", | |
| "_pdf_hash": "", | |
| "abstract": [ | |
| { | |
| "text": "GTE has actively participated in the Third Annual Message Understanding Conference (MUC-3) sponsored b y the Naval Ocean System Center (NOSC) for the past 7 months using its natural language processing system, TI A (Text Interpretation Aid)[5]. During this period, TIA system development consisted of two tasks: (1) lexicon tool development, and (2) tailoring TIA to meet the specific needs of the MUC-3 domain, i .e., terrorism in Lati n American countries. Lexicon tool development facilitates system (lexicon) adaptation to new domains (terrorism , drug interdiction, third world launchings, etc .) by semi-automating the manual task of entering words/phrases into TIA's lexicons. Tailoring the existing TIA system to parse and extract MUC-3 domain messages, has allowed GT E to participate in the MUC-3 conference tasks. The purpose of this paper is to report our techniques and findings during these 7 months of MUC-TIA syste m development. Additionally, GTE's official scores for Phase I and Phase II of MUC-3 will be disclosed and discussed. The remainder of this paper is divided into seven sections describing scoring results, justification and analysis o f scores, system development effort, limiting factors, training, reusability of the resultant MUC-TIA system, and fina l conclusions .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Abstract", | |
| "sec_num": null | |
| } | |
| ], | |
| "body_text": [ | |
| { | |
| "text": "\"Matched Only\" refers to the totals for templates which are matched, i .e., scores are not penalized for missin g or spurious slot fillers (template slot id is an exception to this rule) . \"Matched/Missing\" contains the totals fo r templates which are matched, however scores are penalized for missing, but not spurious, slot fillers. All Templates \" contain totals for templates, however penalizations occur for missing and spurious slot fillers . \"Set Fills Only \" contains the totals for only the slots filled from a finite set .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "1", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The scoring software used during Phase I of MUC-3 has been significantly modified to capture more precis e scoring metrics . Phase I Grand Totals roughly correspond to Phase H's Matched/Missing template scores .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "2", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "discouraging results were confirmed after rescoring tstl-muc3 when recall decreased from 31% for \"Matched Only \" (tstl-muc3) to 28% (tst2-muc3) . \"Matched Missing\" and \"All Templates\" were consistent with scores of 11% fo r tstl-muc3 and tst2-muc3 . This decrease may simply indicate tst2-muc3 is a more difficult message corpus t o understand .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "2", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "One interesting score consistent throughout the entire MUC-3 evaluation task (Phase I and Phase II) wa s precision . Precision (PRE) measures the correctness of the information extracted from the messages and placed i n the templates during the parsing processes . The overall goal is to maximize precision . GTE's precision (for \"tst2 -muc3\") was 43% for \"Matched Only\", 43% for \"Matched/Missing\" and 25% for \"All Templates .\" After examinin g Phase I scores, precision did increase (although not significantly) 1%. Moreover, the rescored \"tstl-muc3\" precision was 42% for \"Matched Only\", 42% for \"Matched/Missing\", and 18% for \"All Templates\" .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Precision", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Overgeneration is the scoring metric which measures extraneous template fills, i .e., the percentage of templates which were incorrectly spawned during the parsing and extraction processes . This metric should be minimized. GTE scored 33% for tst2-muc3 \"Matched Only\", 33% for \"Matched/Missing\", and 61% for \"All Templates\" . Durin g Phase I testing, GTE scored 29% overgeneration . After rescoring tstl-muc3 (after Phase II development) overgeneration increased to 35% for \"Matched Only\", 35% for \"Matched/Missing\", and 72% for \"All Templates\" . Overgeneration slightly increased by Phase II development. 0 0 0 0 0 23 0 100 0 instrument-types 17 0 0 0 0 0 17 0 100 0 incident-location 95 32 8 12 12 0 63 15 44 0 phys-effects 29 0 0 0 0 0 29 0 100 0 human-effects 29 30 2 2 6 20 19 10 10 6 7 GRAND TOTAL 1161 580 216 51 145 168 749 21 42 2 Table 3 .0 : TST1-MUC3 rescored after Phase II developmen t", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 606, | |
| "end": 903, | |
| "text": "0 0 0 0 0 23 0 100 0 instrument-types 17 0 0 0 0 0 17 0 100 0 incident-location 95 32 8 12 12 0 63 15 44 0 phys-effects 29 0 0 0 0 0 29 0 100 0 human-effects 29 30 2 2 6 20 19 10 10 6 7 GRAND TOTAL 1161 580 216 51 145 168 749 21 42 2", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 904, | |
| "end": 911, | |
| "text": "Table 3", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "OverGeneration", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Although the above stated scores seem rather discouraging or low, there are several valid justifications for suc h occurrences. The sections which follow explain each of the justifications.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "JUSTIFICATION AND ANALYSIS OF SCORE S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Phase I scores left GTE with some artificial results for recall and precision . Several slot fillers were the direc t result of system defaults . This in turn filled many slots with correct fillers, but for wrong reasons, a phenomeno n which Grishman calls \"uncoupling input and output .\" For example, during Phase I scoring, the MUC-TIA syste m defaulted the template slot Perpetrator : Confidence to the set list filler of \" REPORTED AS FACT\" ; however, no real analysis was performed . Since \"REPORTED AS FACT\" was the most used correct slot filler, th e score was artificially inflated.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "System Defaults Less", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The MUC-TIA System's internal semantic representation of a parse consists of realizations of structure d concepts . Structured concepts are frame-like knowledge representations which maintain slot fillers . During th e semantic parsing process structured concepts are realized (essentially instantiated) by slot fillers such as simple tex t strings, or with more complex fillers such as demons, which are spawned. For example, an event such as a bombing instantiates a structured concept bombing-p with slots for actor (who performed the bombing), theme (wha t was bombed), location (where the bombing took place), etc . These realized structured concepts in turn represent th e message parsed and maintain the data extracted. A backend translation process then maps and normalizes the dat a maintained in the structured concepts and places it in the appropriate templates .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Backend Translatio n", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "This domain specific back-end translation module has not been fully tested and/or implemented . Man y formatting issues still need to be resolved. Moreover, template merging techniques/heuristics are still being tested to determine optimal methods. Additionally, complete slot cross-referencing has not been completed and fully tested . As a result, many incorrect and partial matches occurred during the scoring process, thereby causing a detrimenta l effect on GTE's scores . Although the correct data was extracted from the message and maintained in the system' s internal representation, i .e., structured concepts, the actual template slot was filled incorrectly due to the back-en d translation process . For example, message TST2-MUC3-0034 ' s HUMAN TARGET : TYPE correct slot filler is : POLITICAL FIGURE : \"JECAR NEGHME\", however, TI A' s response template indicates \" SPOKESMAN\" : \" -\" .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Backend Translatio n", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "After further review of TIA's internal representation of the message, a murder -p structured concept wa s properly instantiated with ,\"JECAR NEGHME\", a SPOKESMAN for the MIR, thereby properly identifying the appropriate human target . GTE has identified these \"data extraction\" problems with the back-end translator and recommends this module be rewritten .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Backend Translatio n", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "During Phase II development several new semantic ideas were implemented which were not fully tested . For instance, to assist in filling the PERPETRATOR : CONFIDENCE slot, a \"mode-p\" prediction prototype [1] was defined which maintains two slots : By-Whom-S, and Insert-Mode-S . The By-Whom-S slot is filled by th e authoritative figure which is found in the last act (this prediction is defined in the mode-p prediction prototype' s control structure .) The \"insert-mode-s\" slot's purpose is to inhibit the generation of a new template . For example , message TST2-MUC3-0011 states Normally, the word rifts spawns a realization of an attack template ; however, the phase, denied that inhibited the attack template . This experimental mechanism has not been fully tested .", | |
| "cite_spans": [ | |
| { | |
| "start": 204, | |
| "end": 207, | |
| "text": "[1]", | |
| "ref_id": "BIBREF0" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "New Semantics Partially Implemente d", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "During the MUC-3 development period, several lexicon tools have been implemented which facilitat e development for new domains, e.g ., terrorism, drug interdiction, third world launches, etc . These semi-automati c tools allow the lexicon developer to browse the message corpus and define lexical entries through a series o f menus 3 . Additionally, sorting utilities were developed which operate on the automatically defined lexical entries . These tools are imperative to training any natural language processing system to a new domain . These tools have greatly increased the lexicon developers productivity while reducing debugging time . Since the majority of MUC -3's development time was devoted towards tool implementation, a minimal amount of MUC-3 domain-specifi c system development was performed, which is reflected in GTE's scores.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Time of Domain Specific System Developmen t", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The majority of the MUC-3 system development effort involved lexicon development issues (discussed below) . The construction of lexicon development tools and macros absorbed the majority of the development time . Approximately 200 of the 360 hours of system development were devoted towards these tasks . The balance , approximately 160 hours, were devoted towards actual MUC-3 task specific system development . As a result, GTE' s scores were adversely affected.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "SYSTEM DEVELOPMENT EFFOR T", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The following sections describe some of the limiting factors and problems which GTE had to overcome in orde r to participate in the MUC-3 Project.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "LIMITING FACTOR S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "This menu approach will be modified and a human machine interface using X11 and Motif will be implemente d for the lexicon development tools in the near future .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "3", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "GTE has devoted two software engineers working on the MUC-3 Project for varied amounts of time . One software engineer (employed by GTE for six years) worked on the original TIA system first established in 1985 . During Phase II development, he devoted approximately 80 hours to MUC-3 domain specific tasks . The other software engineer (employed by GTE for approximately one year) devoted approximately 280 hours towards lexico n tool development, system administration (Sun 4/490 Sparc Server), MUC-3 domain specific system development , scoring and interpreting results. As a result, GTE was not able to consecrate the desired time to MUC-3 (domai n specific) system development .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Person Resource s", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "A second limiting factor which arose and was eventually solved was the syntactic parser's combinatorial explosion problem . This problem occurred due to the top-down exhaustive nature of the parser . The problem originally became apparent when several non-terminal syntactic constituents, e .g., regions, organizations, becam e extremely large and unwieldy . Since the parser expands non-terminals in a uniform, non-heuristic manner, al l applicable grammar rules are fired -even rules which are not viable . For example, if two rules present in the syntactic grammar are of the form :", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Syntactic Parser's Combinatorial Explosion Problem", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Rule 1 :", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Syntactic Parser's Combinatorial Explosion Problem", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "<A> --> <B> <C > Rule 2 :", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Syntactic Parser's Combinatorial Explosion Problem", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "<A> --> <B> <D >", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Syntactic Parser's Combinatorial Explosion Problem", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "and Rule 1 fails because <B> cannot be expanded during the parse, Rule 2 or any other rule of the form : <A> --> <B> was still attempted to be expanded, even though it cannot yield any positive results . Consider the followin g dev-muc3 excerpt (labeled \"Failed String\") and the <Name-Position> syntactic rules shown below . Since the string's parse fails at the non-terminal <Region> in the rust <Name-Position> rule (because president cannot be a <Region>), the parse should not be permitted to try parsing using the second option of <Name -Position> . When the number of nonterminal expansions for a single nonterminal is \"small\", this issue is no t problematic . However, as the number of expansions becomes \"large\", the inefficiency degrades the parse r dramatically.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Syntactic Parser's Combinatorial Explosion Problem", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The problem was solved by establishing/marking the set of non-terminals which may contain a large number o f expansions, and maintaining failed parse states within the current phrases parse . If the current phrase being parsed is in a state which has failed at some prior time and the current nonterminal being expanded is \"large\", the system does not try to expand the current nonterminal using the current rule . This pruning of the search space does not alter th e language recognized, i .e., all previously parsable constructs are still viable and are parsed appropriately . This solution caused dramatic results during several parses . Prior to this optimization, a sample parse of a phrase containing approximately three words which yield \"large\" nonterminals took the MUC-TIA syste m approximately 145 CPU seconds to run . After the optimization was implemented, the same phrase too k approximately 0.4 CPU seconds -obviously a worthwhile improvement .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Syntactic Parser's Combinatorial Explosion Problem", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "As previously discussed, one MUC-TIA training task consisted of automating the process of lexicon development. GTE has developed two tools and several domain specific macros to train the system, each discusse d below in more detail .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "TRAININ G", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The lexicon learner tool/utility automates the process of entering unknown (essentially undefined) words and/or phrases into the appropriate syntactic lexicon with the appropriate syntactic and semantic features . Consider th e following excerpt from one of the dev-muc3 messages.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Lexicon Learner and Sorter Tools", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "\"Ricardo Alfonso Castellar, Mayor of (Achi .UNKNOWN), in the Norther n Department of Bolivar, who was kidnapped on 5 January, apparently by Arm y", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Lexicon Learner and Sorter Tools", | |
| "sec_num": null | |
| } | |
| ], | |
| "back_matter": [ | |
| { | |
| "text": "of National Liberations (ELN) guerillas, was found (slaughtered .UNKNOWN ) today, according to authorities . \"When the lexicon learner encounters the unknown lexical entry \"Achi\", the system prompts for the appropriat e syntactic and semantic information necessary to sufficiently define the lexical entry as shown below. The city Achi is defined by a Def-Region macro which maintains fields for grammar, syntax, part-of, and type . The gramma r field is initialized to mu c 3 (the grammar for the MUC-3 project), s y n t ax (specifies the list of possibl e articulations for the lexical entry) is set to the list consisting of one element, (a chi) , part -o f (specifies th e region's hierarchical constituents) is set to bo 1 iva r, and the type field (specifies the region's demography, e .g . , village, city, state, country, continent, etc .) is set to city . As the defining process continues, the lexicon learner will encounter the second unknown lexical entry \" slaughtered\" , prompt for the appropriate information and then construct the following lexical entry : These two lexical entries are then appended to the appropriate lexicon file and are compiled into the MUC-TI A system during the next Make of the parser.The lexicon learner was run on approximately 750 dev-muc3 messages over a period of approximately on e month . In that time, the MUC-TIA system lexicon grew from approximately 2000 lexical entries to over 25,00 0 lexical entries. During training development all 1200 dev-muc messages could have been run through the lexico n learner. However, due to time constraints and new-word vs. training time return, GTE software engineers elected no t to continue with the lexicon learning .Actual system development took place on approximately 3 dev-muc3 messages . Once again this training statistic occurred due to time and budgetary constraints . GTE plans to continue development in this domain in expectation of participating in MUC-4 next year .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "annex", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The second type of lexicon development facility which was implemented for the MUC-3 project was a series of specialized macros which facilitate the definition of regions, events, people, organizations, terrorist groups, las t names, etc . Each macro performs its unique job by establishing the grammar, syntax, and several macro dependen t specialized fields. For example, clef-region maintains part-of, type, and predicts fields . Moreover, the lexical entry \"slaughtered\" defined above establishes a predicts field of murder-p . This predicts field may trigger an instantiation (not necessarily a realization) of the murder-p structured concept.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Domain Specific Macros", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The majority of the MUC-TIA System may be reusable for other terrorism domains ; however, should an entirely new domain be needed (such as third world launches or aircraft tracking), approximately 75% of the lexico n would need to be replaced . This task is not as insurmountable as it once was (pre-MUC-3) due to the lexicon tool s developed during the MUC-3 project .Additionally, since the back-end translator is very domain specific, a rewrite for the new domain would be necessary to adapt a new template structure .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "REUSABILIT Y", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Although GTE's raw scores seem rather discouraging, we feel significant progress has been made in an effort t o solve the complex natural language/text interpretation problems posed by the MUC-3 Project . Encouragin g factors/developments such as: TIA's speed, expandability, lexicon development tools, and engineering experienc e gained through the MUC-3 effort have positioned GTE in the right direction so that future research and development efforts will succeed not only in the MUC-3 terrorism domain, but in any domain which needs natural language/tex t interpretation technologies. We look forward to participating in next year's MUC-4 .", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "CONCLUSION S", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "GTE would like to thank Beth Sundheim of NOSC for her efforts during MUC-3 . Without her help, many technical and managerial issues/problems would not have been solved . Additionally, GTE would like to thank Charles Wayne of DARPA for providing fmancial support and allowing GTE to participate in MUC-3 .The author would like to thank Robert Dietz of GTE for providing technical support during the MUC-TI A development efforts, and Carl Caley and Ken Duvall of GTE for providing the administrative assistance throughou t the entire MUC-3 effort.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "ACKNOWLEDGEMENT S", | |
| "sec_num": null | |
| } | |
| ], | |
| "bib_entries": { | |
| "BIBREF0": { | |
| "ref_id": "b0", | |
| "title": "Description of the GTE TIA System Used for MUC-3", | |
| "authors": [ | |
| { | |
| "first": "R", | |
| "middle": [], | |
| "last": "Dietz", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1991, | |
| "venue": "Proceedings of the Third Annual Message Understanding Conference (MUC-3)", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Dietz, R ., \"Description of the GTE TIA System Used for MUC-3\", Proceedings of the Third Annual Message Understanding Conference (MUC-3), Naval Ocean Systems Center (NOSC), San Diego, CA, May 21-23, 1991 .", | |
| "links": null | |
| }, | |
| "BIBREF1": { | |
| "ref_id": "b1", | |
| "title": "Software Design Document for Long Range Cruise Missile Analysis and Warning Syste m (LAWS) and Text Interpretation Aid (TIA)", | |
| "authors": [ | |
| { | |
| "first": "R", | |
| "middle": [], | |
| "last": "Dietz", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1990, | |
| "venue": "GTE Government Systems", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Dietz, R., \" Software Design Document for Long Range Cruise Missile Analysis and Warning Syste m (LAWS) and Text Interpretation Aid (TIA)\", GTE Government Systems, August, 1990 .", | |
| "links": null | |
| }, | |
| "BIBREF2": { | |
| "ref_id": "b2", | |
| "title": "The MUC-3 Scoring Software User Manual", | |
| "authors": [ | |
| { | |
| "first": "P", | |
| "middle": [], | |
| "last": "Haverford", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1991, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Haverford, P ., \"The MUC-3 Scoring Software User Manual\", GE Corporate Research and Development , April, 1991 .", | |
| "links": null | |
| }, | |
| "BIBREF3": { | |
| "ref_id": "b3", | |
| "title": "Adapting a Natural Language System for Large, Less-Constrained Domains", | |
| "authors": [ | |
| { | |
| "first": "C", | |
| "middle": [], | |
| "last": "Taylor", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "R", | |
| "middle": [], | |
| "last": "Dietz", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1991, | |
| "venue": "Th e Eighth Annual Advanced Military Intelligence Conference", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Taylor, C ., Dietz, R ., \"Adapting a Natural Language System for Large, Less-Constrained Domains\", Th e Eighth Annual Advanced Military Intelligence Conference, Greenbelt, Maryland, March 12-15, 1991 .", | |
| "links": null | |
| }, | |
| "BIBREF4": { | |
| "ref_id": "b4", | |
| "title": "MUC-3 Phase I at GTE Government Systems", | |
| "authors": [ | |
| { | |
| "first": "C", | |
| "middle": [], | |
| "last": "Taylor", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1991, | |
| "venue": "Advanced Decision Systems (ADS)", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Taylor, C ., \"MUC-3 Phase I at GTE Government Systems \" , Advanced Decision Systems (ADS), Mountai n View, CA, February 12-14, 1991 .", | |
| "links": null | |
| } | |
| }, | |
| "ref_entries": { | |
| "FIGREF0": { | |
| "text": "The chief of the armed forces joint chiefs of staff have categoricall y denied that there are any rifts between Salvadoran army officers an d U .S . military, as asserted by the Washington Post . \"", | |
| "num": null, | |
| "type_str": "figure", | |
| "uris": null | |
| }, | |
| "FIGREF1": { | |
| "text": "Alfredo Cristiani, president of El Salvador \" <Name-Position> --> <Name > <Name> <comma> <Region> <Position> I <comma> <Region> <apostrophe-s><Position>", | |
| "num": null, | |
| "type_str": "figure", | |
| "uris": null | |
| } | |
| } | |
| } | |
| } |