汉语句法分析研究综述  

Research on Chinese parsing

作　　者：郑伟发[1] 

机构地区：[1]广东商学院信息化建设与管理办公室,广州510320

出　　处：《信息技术》2012年第7期72-74,78,共4页Information Technology

基　　金：广东高校优秀青年创新人才培育计划项目资助(LYM09084)

摘　　要：句法分析是自然语言处理的基本问题,掌握句法分析的研究方法和研究现状是进一步深入研究的基础。首先介绍短语结构语法和依存语法的特点,对比基于规则、统计和组块的句法分析方法,并对汉语句法分析的现状和原因进行研究,最后指出汉语句法分析应采用多方法、多特征或多知识源相结合才能高效地进行分析。Parsing is the basal problem of the natural language processing. To master research methods and research status of parsing is the basis of further study. This paper first describes the characteristics of the phrase structure grammar and dependency grammar, and then contrasts to the rule-based parsing method, the statistics-based parsing method and the chunk-based parsing method, and then researches the status of Chinese parsing, and finally points out that Chinese parsing should combine muhi-method, multi-feature knowledge sources in order to efficiently carry out the analysis.

关 键 词：语法体系 汉语句法分析 自然语言处理 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]