基于本体概念相似度的网页排序算法研究  

Research on Ranking Web Pages Based on Ontology Concepts Similarity

作　　者：张健[1] 冯飞[1] 刘宇[1] 马红烨[1] 

机构地区：[1]北京信息科技大学经济管理学院

出　　处：《情报学报》2013年第11期1174-1183,共10页Journal of the China Society for Scientific and Technical Information

基　　金：国家科技支撑计划课题(2012BAH10F01、2011BAC04B02);教育部新世纪优秀人才计划项目(NCET-110893)

摘　　要：网页排序的目的是使主题相关度高的网页排在检索结果靠前的位置，但是由于网络资源规模巨大且呈高度动态的增长，检索结果仍然会存在大量不相关的网页信息。针对这种问题，本文通过分析比较目前对网页排序算法的研究，提出一种基于本体概念相似度计算的网页排序算法，并以盐湖领域为例进行分析。该方法通过构建盐湖领域本体获取本体概念集，并对概念集分类且给予权重，利用概念相似度计算方法计算网页内所有概念与本体概念集中概念的相似度，根据综合得分对网页进行排序。实验结果证明该方法具有可行性和较高的准确率。The purpose of ranking web site is to rank the higher relevance of the web site in the front position of the search results, but due to the huge and highly dynamic growth of web resources, there are a large number of irrelevant web sites in the search results. Aiming at the problem, this paper analyses and compares the current research on web site ranking algorithms, and proposes a method based on ontology concepts similarity computation has been used to rank web site, to salt lake domain as an example. Through building the salt lake domain ontology to obtain the set of ontology concepts, then classifying and giving weights to them, and using concept similarity calculation method to calculate the similarity of all the concepts in the Web site and the concepts among the ontology concept, and according to the comprehensive scores to re-rank the retrieved Web pages. The experimental results show that the new method is feasible and has higher accuracy.

关 键 词：本体 本体概念 网页排序 概念相似度 盐湖领域 

分 类 号：TP393.092[自动化与计算机技术—计算机应用技术]