2000 Fiscal Year Final Research Report Summary
COMPUTER ANALYSYS OF JAPANESE SENTENCES IN OLD WRITING STYLE
| Project/Area Number |
11680422
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
情報システム学(含情報図書館学)
|
|---|
| Research Institution | MUSASHI INSTITUTE OF TECHNOLOGY |
|---|
Principal Investigator |
UEHARA Tetsuzo MUSASHI INST.OF TECHNOLOGY, FACULTY OF ENGINEERING, PROF., 工学部, 教授 (60257102)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
SHIMIZU Yumiko MUSASHI INST.OF TECHNOLOGY, FACULTY OF ENVIRONMENTAL AND INFORMATION STUDIES, ASST.PROF., 環境情報学部, 助教授 (30298020)
ARAI Syuuichi MUSASHI INST.OF TECHNOLOGY, FACULTY OF ENGINEERING, ASST.PROF., 工学部, 助教授 (20212590)
|
|---|
| Project Period (FY) |
1999 – 2000
|
| Keywords | OLD JAPANESE TEXT / CORPUS / JAPANESE LANGUAGE PROCESSING / NATURAL LANGUAGE PROCESSING / GRAMMAR / 日本語処理 |
|---|
| Research Abstract |
We carried out following experiments concerning computer processing of classical Japanese sentences.
1. Dependency Analysys Program with features of external specification of syntactic rules : We designed Dependency Analysys Functions which get syntactic rules of dependency analysis from external files and programmed them and tested on some rules for Japanese classical sentences..
2. Corpus Accessing Tool : We designed and programmed a corpus access tool as a basic set of corpus access functions for EDR-Japanese corpus. Then we designed Sentence Structure Description Language SSDL and programmed its processor which carried out sentence structure collation for EDR Japanese corpus as an application of the access tool. Next, we carried out performance improvement by indexing mechanism on each part-of speech and word spelling.
3. Word Concept and Case Frames for Classical Japanese
We proposed a method of extracting concepts of a classical Japanese word from Classic-Modern Japanese Dictionary. Next we proposed a method of extracting case frames of classical Japanese verbs from the same dictionary.
4. Tagged Corpus of Classical Japanese
We build a corpora with part-of-speech tags and syntax trees on two Japanese classical literature : Ise-monogatari and Taketori-monogatari. They are hand-build.
|
