网络用户兴趣的智能挖掘方法研究  

Research on Intelligent Mining Method for Web Users Interests

作　　者：李培[1] 马力[1] 

机构地区：[1]西安邮电大学,陕西西安710061

出　　处：《计算机技术与发展》2014年第2期76-78,83,共4页Computer Technology and Development

基　　金：国家自然科学基金资助项目(61105064;61203311);陕西省自然科学基础研究计划项目(2011JM8007);西安邮电大学校青年教师科研基金(ZL2013-24)

摘　　要：目前网络上的重要应用都是围绕对用户兴趣的研究和发现而展开和完善的,主要的方式是借助于对用户的Web访问数据进行相关挖掘。该研究主要是通过建立一个从底层数据获取到上层数据处理的原型系统,对真实捕获的网络数据利用小世界网络模型提取中文文档关键字后处理为用户兴趣,再将用户的访问兴趣通过隐马尔可夫模型抽象成一种时间序列,依次反映用户兴趣的序列性,从而利用GSP算法得到用户的兴趣并供后续处理。实验证明,该原型系统从数据获取到最终处理,可以得到比较满意的结果。At present important applications on the network is expanded and improved around research and discovery of user interest, the main way is through Web access to data mining. The research is mainly through the establishment of a prototype from underlying data acquisition to the upper data processing, for an actual network data captured apply the small world network model to extract keyword as user interests, which are extracted a kind of time series by hidden Markov model, which was used to reflect the sequential features of the user interest. Use asp algorithm to mine the user＇s interests for processing. Experiments show that the prototype system from data acquisition to final disposal can be more satisfied with the results.

关 键 词：兴趣挖掘 文本聚类 智能算法 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]