摘要
为了改进基于关键词的信息检索方法的局限性，论文研究了一种综合利用领域本体改善信息检索性能的方法。该方法强调通过交互式的方式引导用户一步步逼近其真实的、潜在的检索需求，使用基于编辑距离的词形匹配方法辅助用户查询本体词汇，使用基于概念空间的检索词联想方法帮助用户扩充检索词。使用基于领域本体的词义识别算法来确定文档中的词汇词义。使用XML技术实现用户查询需求和文档标注的规范化标注。实验表明，该方法会有效提升查全率并且会改进查准率。
In order to improve the performances of keywords-based information retrieval method, a method of using field ontology in in,rmation retrieval is introduced. It focuses on guiding user to find their actual and potential search requirements step by step which depends on two methods, one is using edit distance based method to search word in ontology base and the other is using concept space based method to help user to extend key words. Moreover it utilizes the relations between words built by filed ontology to find word＇s actual meanings in it＇s context, and finally one XML scheme is described simply which is used to index key words selected by user and processed document . Experiment results show that this method can improve precision ratio to a certain extent, but recall ratio greatly.
出处
《情报学报》
CSSCI
北大核心
2010年第2期215-222,共8页
Journal of the China Society for Scientific and Technical Information
基金
南京航空航天大学引进人才基金（1009-234039）
某国防技术基础项目.
关键词
信息检索
领域本体
概念空间
information retrieval, ontology, concept space
