基于用户浏览行为的用户兴趣模型的表示及更新

Representation and Update for User Interest Model Based on User Browsing Behaviors

在分析用户浏览行为获取用户兴趣的基础上,提出一种树状结构模型与向量空间模型相结合来表示用户兴趣模型的方法,基于此用户兴趣模型,把遗忘机制引入模型的更新。这种兴趣模型不但具有加权矢量模型的加权关键词的特点和层次模型的层次性,并且还结合向量空间模型的特点,将用户的兴趣与浏览页面的特征向量结合起来。实验证明,建立用户兴趣模型的方法是合理和有效的。

Based on analysis of user browsing behaviors to extracting user interests,a method of user model representation of tree struc-ture model combined with vector space model.Based on this representation of user model,a forgotten mechanism is intro-duced to update of this model.This model not only has characteristics of weighing vector model and keywords,and hierarchy of the hierarchical model,but also combines with characteristics of vector space model to make user interests and eigenvector of browsing pages consistent.Experiments results prove that the researching method of building user interest model is reason-able and effective.