支持Internet上个性化信息重组与发布的Web挖掘关键技术的研究

Internet上个性化信息的重组与发布是Web个性化技术的一个重要组成部分,这一领域目前存在的主要问题是:并非没有信息重组和发布的工具,而是缺乏能够使这类工具高效工作的支持技术。本文提出一种将流数据处理技术引入Web点击流、IP地址流及页面文本流挖掘和分析过程,研究基于Web数据流挖掘的用户行为和需求分析方法;将本体和领域知识引入Web内容挖掘过程,研究领域知识指导下的Web内容挖掘方法;将基于Web数据流挖掘的用户行为和需求分析与领域知识指导下的Web内容挖掘相结合,研究Internet上Web信息模式和Web用户模型及其相互关系的建立;将上述研究成果应用于实际,以期达到高效地支持Internet上满足用户个性化要求的信息重组与发布的目的。

Study on Key Techniques of Web Mining for Personalized Information Reorganization and Distribution

The personalized information reorganization and distribution is an important part on Web personalization. Now the major problem of this domain is not lacking the tools for information reorganization and distribution, but re- quirinig the supporting techniques for them work effectively. This paper proposes an approach including: Stream data processing technique is imported in analyzing Web click stream, IP stream and text steam for studying users' behavior and requirement based on stream data mining. Ontology and domain knowledge is introduced in Web content mining for guiding the mining process. Users'behavior and requirement analysis based on Web stream data mininig is com- bined with Web content mininig under the guidance of domain knowledge for creating Web information pattern, user model and their relation. The study results are used for supporting effectively personalized information reorganization and distribution of Internet according to users'requirement.