上海大学学报(自然科学版) ›› 2010, Vol. 16 ›› Issue (3): 318-322.

• 计算机工程与科学 • 上一篇    下一篇

基于特定群体兴趣的混合个性化推荐算法

毕行,徐炜民   

  1. (上海大学 计算机工程与科学学院,上海 200072)
  • 出版日期:2010-06-28 发布日期:2010-06-28

A Hybrid Personal Recommendation Algorithm Based on Designated Group Interest

BI Hang,XU Wei-in   

  1. (School of Computer Engineering and Science, Shanghai University, Shanghai 200072, China)
  • Online:2010-06-28 Published:2010-06-28

摘要:

个性化推荐服务能够为网络用户提供针对兴趣偏好的推荐项目资源,现已被成熟地运用到网站导航、数字化图书馆检索系统、电子商务以及搜索引擎等领域.在研究有关推荐技术以及混合方式后,提出一种基于特定群体的混合推荐算法,其紧密结合了模糊聚类与两种协同过滤技术.实验结果表明,该算法不仅有效地解决了数据集的稀疏性问题,而且在一定程度上改善了推荐结果的质量.

关键词: 个性化推荐;特定群体;模糊聚类;协同过滤

Abstract:

Personalized recommender services provide users with the preference items based on their interests.It has been applied to various areas such as website navigation, indexing and retrieving system of digital libraries, Ecommerce, search engine. Having studied recent developments in the relavent areas, we propose a new algorithm based on a esignated group that using fuzzy clustering combined with two collaborative filtering techniques. The result shows that the roposed algorithm can solve the problem of sparsity and improve quality of preferences.

Key words: personal recommendation; designated group; fuzzy clustering; collaborative filtering

中图分类号: