| 摘要: |
| 基于蒙太古语义学(MS)和受控自然语言(CNL)的文献信息系统,兼具传统文献检索分类法的“符号化”和主题法的“规范词”的特点,并带有复杂的语法和语义功能;从而可以大大提高文献信息的检索和利用效率,还能自动生成规范的文献摘要,并完成不同语言之间的自动翻译.尝试实现这种文献信息系统的结构,其中包括构建语词的范畴分类和语句的函数表达式,以及自然语言-形式语言、中文-英文之间的映射(翻译)等;最后设计一个应用程序以验证该系统的若干功能.本研究表明:这种全新概念的信息组织和管理系统是可行的. |
| 关键词: 蒙太古语义学 受控自然语言 机器翻译 文献信息检索 |
| DOI： |
| 分类号: |
| 基金项目:上海哲学社会科学研究基金(2012BZX003);上海城市管理职业技术学院研究课题(CGZW1401) |
|
| Implementation of document information system based on Montague semantics and controlled natural languages |
|
CAO Yu1, WANG Shanping2, ZHU Shuilin3
|
|
1.Shanghai Technical College of Urban Management,Branch of Humanities and Information Technology;2.Editorial Office of Journal ECNU,East China Normal University;3.Institute of Information,Shanghai Academy of Social Sciences
|
| Abstract: |
| A document information system based on Montague semantics(MS)and controlled natural languages(CNL)can be both symbolized and normalized-the two characteristics possessed by traditional document classification systems and thesaurus indexing systems,respectively;it can also realize complicated syntax and semantics functions,so as to improve the efficiency of document information retrievals greatly.Moreover,it can form standard paper abstracts automatically and realize translations between different languages.This paper attempted to construct such a kind of system,which includes category classification of words,expressions of sentence functions,maps/translations between natural and formal languages,as well as between Chinese and English languages.Finally,an application program was designed to test functions of the system.The results of this study show that the data structure and function simplementation of the proposed information system based on new concepts are feasible. |
| Key words: Montague semantics controlled natural languages machine translation document information retrieval |
