摘要
目前的搜索引擎普遍存在着查全率和查准率不高的问题,任何一个简单的查询都可能返回数以万计的检索结果,而其中只有很少一部分与用户真正的检索要求有关,对查询的处理是基于本体(Ontology)的语义检索最重要的部分.针对现有查询分析方法的不足,提出了一种基于Ontology的综合词义关系和语义关联分析的查询分析算法,给出了基于本体映射的语义相似度算法,通过对用户输入关键字词义特性和本体实例之间语义关联强弱的分析,提高了用户输入关键字到本体概念映射的完整性和准确率,保证了用户查询和检索语言在语义上的一致性,提升了查询的满意度.
Now search engines generally have the problem of lower recall and precision. Any simple query may return thousands of results for retrieval, but few of results relate with the genuine requirement for the users. It is the most part of semantic retrieval based on ontology to deal with the queries. Aiming at the shortage for existing methods in analysis of query, this paper presents an arithmetic with analysis of query which integrates the acceptation relation and semantic cor- relative analysis, proposes an arithmetic based on calculation of similarity, analyses the acceptation characteristic of key- words and the acceptation correlative intensity between ontology instances that users input into the system, improves the recall and precision of the mapping from keywords inputted by users to ontology concept, ensures the semantic coherence between query and retrieval language, and promotes the satisfaction for querying.
出处
《南京师范大学学报（工程技术版）》
CAS
2008年第4期63-67,共5页
Journal of Nanjing Normal University(Engineering and Technology Edition)
关键词
本体
查询分析
语义
信息检索
ontology, analysis of query, semantic, information retrieval
