基于协同推荐的web日志预处理过程

个性化推荐技术是电子商务系统中重要的技术,但对一般的非商务型网站如何向用户提供推荐服务成为当前研究的热点。Web日志记录了用户访问网站的详细信息,这为推荐技术提供了新的研究领域。本文提出了针对协同推荐算法的web日志预处理全过程。并对预处理过程的用户识别、会话识别、路径补充、用户兴趣评估进行了详细的探讨并提出了自己的见解。

Web Log Preprocessing Based On Collaborative Filtering Algorithm

Personalization recommendation technique is an important technology in E_commerce, but how to provide recommendation service in net site which is not E_commerce is becoming a hot research now. Web log records user-behavior in detail, this provides a new research field for recommendation technology. This paper introduces the traditional process of collaborative filtering and proposes a web log preprocess on collaborative filtering algorithm, This paper also proposes new ideas in user-identification and user-interest-rating.