Trial Construction of an Advanced Computer-readable Lexical Database Capable of Automatic Acquisition of Lexical Information

自动获取词汇信息的先进计算机可读词汇数据库的试建

基本信息

批准号：
07558274
负责人：
FUJISAKI Hiroya
金额：
$ 1.34万
依托单位：
The Science University of Tokyo
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (A)
财政年份：
1995
资助国家：
日本
起止时间：
1995 至 1996
项目状态：
已结题

项目摘要

(1) Expansion of lexical data and classification of newspaper article data : Existing lexical data have been expanded using "Shin-Meikai Kokugo Jiten" (by Sanseido Publishing Co.) and EDR Electronic Dictionary (by Japan Electronic Dictionary Research Institute, Ltd.), and newspaper article data have been classified.(2) Determination of data structure for describing the semantic system : The data structure has been determined on the basis of the EDR Electronic Dictionary, and the data for the semantic system have been classified on a computer.(3) Design and implementation of the subsystem for automatic detection of unknown words : A program has been designed and implemented for the morphological and syntactic analysis of text, and for detecting unknown words.(4) Design and implementation of the subsystem for automatic inference on syntactic and semantic information of unknown words : The data structure has been determined and the program has been designed and implemented for the subsystem for automatic inference of systematic and semantic information of unknown words.(5) Implementation of the basic part of the advanced lexical database system and preliminary confirmation of its operation.(6) Determination of the detailed specifications for the lexical database system : A 40M byte main memory was adopted for the system.(7) Determination of the detailed specifications for the lexical data : The total numbern of lexical items are : 187,868 nouns, 645 pronouns, 10,620 verbs, 1,124 adjectives, 1,345 adverbs, and 144 others.(8) Construction of the advanced lexical database system : The system has been constructed using Arity/Prolog language, and currently occupies 142k bytes of memory.(9) Evaluation of the system : The performance of the system has been evaluated using the lexical data and the electronic texts from newspaper articles, and the results congirmed the basic validity of the current system.

(1) 词汇数据的扩展和报纸文章数据的分类：使用“新明海国语事典”（三精堂出版社）和 EDR 电子词典（日本电子词典研究所）对现有词汇数据进行了扩展(2)确定描述语义系统的数据结构：在EDR电子词典的基础上确定了数据结构，确定了语义系统的数据(3)生词自动检测子系统的设计与实现：设计并实现了文本的词法分析、句法分析以及生词检测的程序。(4)生词自动检测子系统的设计与实现。生词句法语义信息自动推理子系统的实现：生词系统语义信息自动推理子系统的数据结构已确定，程序已设计并实现。 (5) 生词系统语义信息自动推理子系统的实现进阶基础部分词汇数据库系统及其运行的初步确认。(6)词汇数据库系统详细规格的确定：系统采用40M字节主存储器。(7)词汇数据详细规格的确定：总计词汇数量为：187,868 个名词、645 个代词、10,620 个动词、1,124 个形容词、副词1,345个，其他144个。(8)高级词汇数据库系统构建：系统采用Arity/Prolog语言构建，目前占用内存142k字节。(9)系统评价：系统性能使用词汇数据和报纸文章的电子文本进行了评估，结果证实了当前系统的基本有效性。