中文摘要：

针对当前Web挖掘环境下个性化服务的性能不够高效的问题，运用协同过滤技术的理论与方法。研究了一种基于协同过滤的推荐阀值方法，并且提出了一种在线推荐模型。实验表明，在提高Web个性化服务方面，该模型具有更高的效率。

英文摘要：

Collaborative filtering （CF） is a popular technology for building personalized service system. To improve personalized service performance under web-mining environment, a method of recommending threshold based on collaborative filtering is presented and a new online personalized recommendation model is proposed. Experimental results show this recommendation model is more effeetive in improving web personalized service.