针对学术定义的抽取规则构建方法研究  

Studies on Methods of Formulating Rules for Academic Definition Extraction

作　　者：化柏林[1] 刘一宁[1] 郑彦宁[1] 

机构地区：[1]中国科学技术信息研究所,北京100038

出　　处：《情报理论与实践》2011年第12期5-9,共5页Information Studies:Theory & Application

基　　金：国家自然科学基金项目"基于句子匹配分析的知识抽取研究与实现"的研究成果之一;项目编号:70803048

摘　　要：学术定义是学术研究中一种基础性的知识。科技文献中蕴含着丰富的定义,对定义进行自动抽取是知识抽取的一种应用。这些定义具有一定的特征与规律,利用这些特征与规律形成规则对定义进行抽取是常见的方法。定义抽取中最重要的工作是规则构建。文章通过对定义语句的结构方式和分类进行研究,提出了一种基于多重规则的学术定义抽取规则构建方案。该方案通过构建模式规则、句法规则以及加权词规则,构建学术定义抽取所需要的多重规则。Academic definition is a kind of basic knowledge in academic study. Academic documents contain a wealth of definitions. Automatic extraction of definition is a kind of application of knowledge extraction. The definitions have certain characteristics and regularities. Using these characteristics and regularities to form rules to extract definitions is a common method. The most important work in definition extraction is formulating rules. Based on the study of the structure and classification of the definition of sentence, this paper proposes a scheme of formulating rules for academic definition extraction based on multiple rules. By formulating pattern rules, grammar rules and weighted word rules, the scheme establishes the multiple rules required by academic definition extraction.

关 键 词：学术定义 知识抽取 规则构建 

分 类 号：G350[文化科学—情报学]