摘要
传统基于关键字匹配检索策略由于关键字的机械字符匹配和一词多义问题很容易造成漏检和错检。为此,从领域本体概念入手,结合关键字检索的特点,提出一种基于关键词语义扩展的检索策略。策略使用领域本体描述的语义结构扩展关键字匹配范围,避免完全机械的字符匹配造成的漏检,从而提高检索的查全率。在此基础上,利用领域本体中语义相关度过滤检索无关结果,以提高检索的查准率,并根据检索结果与本体语义相关度算法排序。
Traditional retrieval strategy based on keyword-matching is facing unavoidable bottleneck because of mechanical characters matching and polysemy problems. A retrieval strategy based on domain ontology and semantic expansion of keywords was proposed. On the one hand, the domain ontology concept structure was used to expand the retrieval area; on the other hand, a method was presented to filter the irrelevant results. In this way, both the recall ratio and the precision ratio could be improved simultaneously. After that, an algorithm of sorting the retrieval results by the similarity between ontology and result was given.
出处
《计算机应用》
CSCD
北大核心
2009年第6期1575-1577,共3页
journal of Computer Applications
基金
国家973计划项目(2008CB317110)
关键词
领域本体
扩展
查全率
查准率
domain ontology
expansion
recall ratio
precision ratio