一种针对特定领域的查询扩展方法  

A QUERY EXPANSION METHOD AIMING AT SPECIFIC DOMAIN

作　　者：韩仙玉[1] 闫东亮[2] 李睿[1] 

机构地区：[1]阿坝师范高等专科学校网管中心,四川汶川623002 [2]东北大学信息科学与工程学院,辽宁沈阳110004

出　　处：《计算机应用与软件》2014年第1期26-29,共4页Computer Applications and Software

基　　金：四川省教育厅项目(13ZB0037);阿坝师范高等专科学校规划项目(ASB12-27)

摘　　要：为提高特定领域信息检索的效率,针对用户的查询需求不能准确、清晰地表述查询意图的问题,提出一种基于领域本体的语义查询扩展算法。该算法结合领域本体中知识库和实例库的特点,设计相应的扩展算法。其中,基于知识库的扩展主要依赖于等价关系、父类关系和子类关系等层次结构关系。基于实例库的扩展方法主要有:数值型属性扩展和基于统计的语义字典的文本型属性扩展。最后,利用实验验证语义扩展算法的有效性和可行性,并与相关研究方法进行比较,表明该方法在查全率、查准率上也取得了较好的效果。For improving the retrieval efficiency of domain-specific information, we propose a domain ontology-based semantic query expansion algorithm in light of the problems that users query requirement can not express their query intentions accurately and clearly. The algorithm combines the features of knowledge base and instance base in domain ontology to design correspondent expansion algorithm. In it the expansion based on knowledge base mainly relies on the hierarchical relationship including equivalence relationship, super class relationship and subclass relationship; the expanding approach based on instance base mainly includes the numerical attribute expansion and the semantic dictionary text-type attribute expansion based on statistics. Finally, we use experiment to verify the effectiveness and feasibility of the semantics expansion algorithm, and compare it with correlated study methods. It indicates that this algorithm achieves better effect in recall and precision.

关 键 词：查询扩展 语义信息 领域本体查询需求 查询效率 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]