综合用户特征和项目属性的协作过滤推荐算法  

Collaborative filtering recommendation algorithm based on user characteristics and item attributes

作　　者：孙龙菲 黄梦醒[1] 

机构地区：[1]海南大学信息科学技术学院,海口570228

出　　处：《计算机应用研究》2014年第2期384-387,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(71161007);国家教育部社科基金资助项目(10YJCZH049);海南省自然科学基金资助项目(612132);海南省重点科技计划资助项目(ZDXM20120061)

摘　　要：通过分析传统协作过滤推荐算法面临的数据集稀疏性问题及当前解决方法的优缺点,在基于项目的协作过滤推荐算法的基础上,提出了一种综合用户特征和项目属性的协作过滤推荐算法。通过分析不同特征的用户对项目的各种属性的兴趣度,综合已评分的项目属性预测未评分项目,降低数据集的稀疏性,提高项目相似度计算的准确性。在MovieLens数据集上的实验结果表明,在数据极端稀疏的情况下,能够有效地降低数据集稀疏性,并在一定程度上缓解了协作过滤推荐算法中的冷启动问题,提高了推荐算法的预测准确度。Abstract： After analyzing the sparsity of traditiona｜ collaborative filtering algorithm and the merit and demerit of current solu- tions, this paper proposed a collaborative filtering recommendation algorithm based on user characteristics and item attributes which was mainly based on the item-based collaborative algorithm. It predicted the unrated items by analyzing different users＇ interests to various attributes of items and integrating the attributes of rated items to reduce the sparsity of data sets, and then to improve the accuracy of items＇ similarity calculation. The experimental results based on MovieLens data set show that the sparsity of extreme data sets can be reduced effectively, to a certain extent the proposed algorithm alleviates the cold starting problem in collaborative filtering algorithm and achieves better prediction accuracy.

关 键 词：协作过滤 稀疏性 用户特征 项目属性 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TP301.6[自动化与计算机技术—计算机科学与技术]