基于主题图的用户兴趣挖掘模型研究

摘要

User profiles is core of the personalized services. Mining of user interests can not only discover thepotential knowledge, but offer more optimized services. In this paper, we introduce topic map into this research, and study the representation of user interests knowledge using topic map, and based on which, deeper analysis are carried out on the topic map constructed using the scale-free graph clustering algorithm, in the end builds an user interests mining model base on topic map. We deeply analyzed the four key problems besides explaining every function module including scale-free graph clustering algrithm, topic map representation in documents and topic generalization, topic map merging, and construction of intelligent topic map, realizing the process modeling and things modeling.%用户兴趣模型是个性化服务的核心,对用户兴趣的挖掘可以发现潜在的兴趣知识,提供更为优化的服务.本文将主题图技术与用户兴趣模型结合起来,研究了用户兴趣知识的主题图表示,并在此基础上运用无尺度图K-中心点聚类算法对构建的主题图进行深层次的聚类挖掘,建立了基于主题图的用户兴趣挖掘模型.在解释模型各个模块功能的同时,提出了该过程模型中的关键问题,并对建立模型过程中的无尺度图K-中心点聚类算法、文档中的主题图表示及主题概化和主题图合并等关键问题进行了深入的分析,最终构建了智能主题图,实现了过程建模和事物建模.