基于中文信息MMT模型的句法自动分析

Syntactic Automatic Analysis Based on Chinese Information MMT Model

摘要: 本文叙述采用中文信息MMT模型对句子进行句法自动分析的工程实践。研制了可在互联网在线运行的句法分析专家系统。对使用的理论、方法和工具作了详细介绍。介绍了粒计算与符号学理论在句法分析中的应用。研究了句法分析专家系统的组成结构、实施方式和系统设计。最后列举了句子分析实例。在研究方法的决策中采用的是基于理性主义的规则方法。运用代数语言学的理论构建句法分析器专家系统，在研究中使用了中文信息MMT模型，研究结果表明，基于中文MMT模型是进一步解决处理当前句法分析问题的可借鉴的手段，减少了歧义和分析层次。其研究成果的意义在于可以检验语法规则，可以通过语料库的深加工而建立句子的语法模型结构，为深层次的自然语言理解，扩充语法知识库，为句法语义的深入研究提供实用的工具和模型。

Abstract: This paper describes the engineering practice of syntactic automatic analysis of sentences using the Chinese information MMT model. A syntactic analysis expert system that can run online on the Internet is developed. The theories, methods and tools used are introduced in detail. The applica-tion of granular computing and semiotics in syntactic analysis is introduced. The structure, im-plementation mode and system design of expert system for syntax analysis are studied. Finally, examples of sentence analysis are listed. The rationalism based rule method is adopted in the de-cision-making of research methods. The expert system of syntactic parser is constructed with the theory of algebraic linguistics, and the MMT model of Chinese information is used in the research. The results show that the MMT model based on Chinese is a useful tool to solve the current syntactic analysis problems, which reduced ambiguity and analysis level. The significance of the research results is that it can test grammatical rules, establish sentence grammatical model structure through deep processing of corpus, and expand grammatical knowledge base for deep under-standing of natural language. It provides practical tools and models for in-depth study of syntactic and semantic meaning.