基于协同过滤的推荐系统

本文先对网站日志文件进行研究并预处理，将用户访问网站形成的日志文件转换成用户访问模式；接着用对用户访问模式进行K-mean聚类分析，提取具有代表性的用户访问模式；最后用协同过滤推荐技术向网站访问者进行推荐，提供个性化服务，从而实现网站的个性化推荐。

This article firstly researching and preprocessing the original web log data to website visit pattern. Then, we use the K-mean Clustering algorithm to cluster the website visit patterns, finally, we use collaborative filteringbased recommendation technology to provide the website visitors the personalized services so as to generate personalized recommendation.