基于Ontology的面向主题的网页关系识别  

Recognition of the Topic-oriented Web Page Relations Based on Ontology

作　　者：樊闻斌[1] 王守芳[1] 金浩[1] 潘金贵[1] 

机构地区：[1]南京大学计算机软件新技术国家重点实验室,江苏南京210093

出　　处：《华南理工大学学报（自然科学版）》2004年第z1期37-41,共5页Journal of South China University of Technology(Natural Science Edition)

摘　　要：对万维网上同一主题下的大量网页之间进行特定语义关系的分析,并将其应用于搜索引擎中,可以实现智能化的查询和提供其他个性化服务.本研究借助于Ontology中的关系实例,在网页分类的基础上对网页之间的关系进行自动识别,同时提出了网页关系识别规则的自动生成和优化方法.将网页关系识别过程应用于Dolphin系统中,使用户能够在系统的指导下,快速地定位到与起始页面有一定语义关系的其他类型页面中,大大节省了浏览和寻找的开销.The analysis of the semantic relations among large amount of Web pages with the same topic in WWW can realize the intelligent query and provide characteristic services in search engines. In this paper, according to the relation entities in Ontology, the automatic recognition of the relations among Web pages is implemented based on the Web page classification. Also, the automadc generation and optimization method for the relation recognition rules is proposed. It is found that the application of the recognition process to Dolphin system ensures users to rapidly find the Web pages, which have a semantic relation with the initiative page, thus saving the cost of browsing Web pages and finding a suitable one.

关 键 词：ONTOLOGY 技术 网页关系识别 万维网 内容挖掘 

分 类 号：TP393.09[自动化与计算机技术—计算机应用技术]