项目名称: 面向语义检索的汉语名名组合自动释义研究
项目编号: No.61300152
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 自动化技术、计算机技术
项目作者: 王萌
作者单位: 江南大学
项目金额: 22万元
中文摘要: 名名组合是最简洁的结构表达式,在英语和汉语中产量极高,具有衍生能力强、组成方式简单但是歧义性高的特点。名名组合内部隐含着复杂的意义构建过程,存在语义压缩,常常隐含了谓词。对名名组合进行释义的主要目的是发现两个名词之间隐含的谓词,进而揭示这两个名词之间的语义关系。其研究成果不但能够服务于信息检索、问答系统等自然语言处理系统的需求,而且可以为研究心理语言学关于概念组合的认知机制,提供语义解释的依据。本课题的重要创新是采用动态策略,用基于动词的方法研究汉语名名组合的语义解释,探索名名组合隐含谓词的获取方法,并对其语义组合模式进行逻辑建模,建立在计算机上操作的理解模型。本课题将分析结果应用于网络搜索词中的名名组合,实现语义计算在搜索引擎中的初步应用。
中文关键词: 名名组合;自动释义;语义模版;隐含谓词;
英文摘要: Noun compound, the most concise structure which is prolific, simple and ambiguous, often appears in both Chinese and English. There is semantic compression in the complex meaning construction of noun compounds.A predicate is often implied by the Chinese noun compound, and the aim of noun compound interpretation is to recover this implying predicate to reveal the semantic relations between the two nouns. Our research will not only serve the needs of natural language processing systems for information retrieval and answering system, but also provide semantic interpretation for the study of cognitive mechanisms of concepts combination in psycholinguistics. The main innovation in this research is the use of dynamic strategy in which we explore the acquisition method of implying predicate, and establish the logical models for the semantic combination patterns. We also analyze the noun compounds in searching words, and conduct preliminary applications of semantic computing in the search engine.
英文关键词: Noun compounds;automatic interpretaion;semantic pattern;implying predicate;