用户兴趣模型在个性化营销中的研究

摘要

采用隐式方法获取用户信息，创建用户兴趣描述文件。通过用户浏览页面集的内容信息和行为信息以及根据用户在初始化注册信息中提供的兴趣爱好信息进行兴趣度分析。加入空间向量区别地域性兴趣差异，及时准确反馈用户兴趣。采用k－means聚类算法进行聚类分析，得到用兴趣分类树表示的用户兴趣模型。%This paper introduces a method about create user interest description file,and analyzes the interestingness through browse web and user's hobby from register and combine the behavior.Add spatial vector to distinguish geographic difference.This way can feedback user information in timely.This paper takes k-means clustering algorithm to get the interest classification tree representation of user interest model.