中文摘要：

针对传统的协同过滤算法在电子商务系统中存在数据稀疏性和扩展性方面的问题,提出了一种混合用户和项目协同过滤的电子商务个性化推荐算法。该算法采用聚类技术,将基于用户协同过滤和基于项目的协同过滤结合起来进行双重聚类,结合基于用户协同过滤和基于项目协同过滤两方面的优点,从而获得更好的性能。实验表明,通过与其他推荐算法的比较,文中算法具有较高的推荐质量,更好的准确率和召回率。

英文摘要：

In view of the traditional collaborative filtering algorithm in E-Commerce system data sparseness and scalability issues, a hybrid user and item based personalized collaborative filtering recommender algorithm in E-Commerce was proposed. Combined with user based collaborative filtering and item based collaborative filtering, the algorithm uses the clustering technology to cluster twice, and can get better performance. Experiments results show that the algorithm is superior to other recommendation algorithms obviously in the aspect of recommender quality, precision and recall rate.