| Project/Area Number |
60880008
|
|---|
| Research Category |
Grant-in-Aid for Developmental Scientific Research
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Informatics
|
|---|
| Research Institution | Kyoto University |
|---|
Principal Investigator |
TSUJII Jun-ichi Kyoto Uniersity, Faculty of Enginerring Assouinte Prof., 工学部, 助教授 (20026313)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
MURATA Ken-ichi Information Procesing Assouition, Reseacher, 研究員
YAMANASHI Masaaki Kyoto University, Assouite Prof, 教養部, 助教授 (80107086)
NAKAMURA Jun-ichi Kyoto University, Assistant Prof, 工学部, 助手 (30164304)
NAGAO Makoto Kyoto University, Faculty of Enginering Prof, 工学部, 教授 (30025960)
|
|---|
| Project Period (FY) |
1985 – 1987
|
| Project Status |
Completed (Fiscal Year 1987)
|
| Budget Amount *help |
¥30,300,000 (Direct Cost: ¥30,300,000)
Fiscal Year 1987: ¥4,100,000 (Direct Cost: ¥4,100,000)
Fiscal Year 1986: ¥12,600,000 (Direct Cost: ¥12,600,000)
Fiscal Year 1985: ¥13,600,000 (Direct Cost: ¥13,600,000)
|
|---|
| Keywords | Machine Dictioray / Dictionary Data Base / General Purpose Dictionary / Natural dangvase Processing / Machine Translation / 情報検索 / 意味記述 / 自然言語理解 / 辞書 / データベース / 日本語処理 |
|---|
| Research Abstract |
The main objectives of the research are to develop a large dictinary data base which can be used for various fields of natural language processing by computer, and to clarify technical problems in maintaining such a large dictionary data base system. The followings are the main research results we obtained.
(1) Data Extraction from Hyman Dictionaties: Useful lnformation which is contained in longman Contrmporary English Dicitonary was extracted by the algorithm we developed and is managed by a relational data base. The semantic relationships among words are also extracted by analyzing the definition parts of dictionary descriptions. The extracted semantic relationships form a kind of semantic network. Relationships such as IS-A, PART-OF and so on are extracted for not only nouns but also verbs.
(2) Flexible Retrieval Systems: Two retieval systems are developed for managing the dictinary data base in(1 ). One is implemented by using commercially abailable RDB and the other one is implemnted on a PROLOG mac hine. The latter system facilitates flexible retreivals by using inderence capability of PROLOG. Several problems in the managemant of large dictionary data bases have also been revealed.
(3) Morphological Nalysis Program: In order to demonstrate the effectiveness of the data hase, a morphologic al analysis program for English is developed and applied to diversified sorts of texts. More then 90 % of words are correctly recognized by the program, which shows the coverage of the dictionary data bese is suffieient for ordinary NLP applications.
(4) A reports is published in which desctiptions of data formats of dictionary data base and a comprehensive list of extracted semantic relationships are included.
|
