信息检索技术中基于语义的扩展查询研究  

Research on the Semantic Query Expansion of Information Retrieval Technology

作　　者：李兴春[1] 

机构地区：[1]重庆文理学院教学部,重庆永川402160

出　　处：《重庆师范大学学报（自然科学版）》2013年第4期113-116,共4页Journal of Chongqing Normal University:Natural Science

基　　金：重庆市自然科学基金(No.CSTC2011JJA40011);重庆市教委科技项目(No.KJ111216)

摘　　要：用户查询与文档之间语义匹配但词法不匹配现象是影响信息检索效果的重要原因之一。鉴于语义检索受限于本体自身的质量,为了降低其对检索效果的影响,通过分析目前语义查询扩展的研究现状,在已有概念相似度计算算法研究基础上进行改进和融合,提出了一种基于本体的信息检索查询扩展方法,并主要对基于本体技术的概念相似度计算算法进行修正,得到了组合向量空间模型QCR(Q,Ci)=∑k=1,…,Kwk*Sim_Rel(qK,Ci),作为引入查询扩展后的查询结果相关度评价方法。这种方法中,通过建立本体模型并计算本体中概念间的语义相似度来确定扩展查询词,它可以根据用户输入的名称,检索出相关文档并由用户自由设置相似度阈值,并将普通主题检索与语义检索合并,在本体乏力时返回普通检索结果,这在一定程度上弥补了垂直检索系统发展的不足。One of the most important reason to affect the information retrieval result is the phenomenon of semantic match of the user query and the file while syntactic mismatch.Because the semantic retrieval is limited by its quality,to decrease its influence on the retrieval result,through the analysis of the present situation of the semantic query expansion study,based on the improvement and integration of the algorithm study of the similarity of present concept,a new semantic query expansion of information retrieval technology was put forward,in which the integration was revised,and the mode QCR（Q,Ci）=∑ k=1,…,Kwk＊Sim_Rel（qK,Ci） was obtained and used as the evaluation method.In this method,by building the model and calculating the similarity of the concept similarity to determine the query word,it can search the related files and set up the threshold value by the user based on the query word,and the common subject retrieval and semantic query are combined,in which common retrieval results were returned when the noumenon couldn＇t give the results,which completed the development of vertical search system to some degree.

关 键 词：信息检索 语义相似度 扩展查询 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]