一种个性化协同过滤混合推荐算法

A Kind of Personalized Collaborative Filtering Hybrid Recommendation Algorithm

传统协同过滤算法主要根据稀疏的评分矩阵向用户作出推荐,存在推荐质量较差的问题。为此,提出一种基于信息熵的综合项目相似度度量方法。考虑到用户的兴趣会随时间发生变化,而且不同用户群体的兴趣变化不同,受艾宾浩斯记忆遗忘规律启发,提出适应于不同用户群体兴趣变化的数据权重。基于movielens数据集的实验结果表明,改进后算法不仅能缓解评分数据稀疏问题,而且能提高算法的准确率。

Traditional collaborative filtering algorithm exists poor recommendation quality for recommending to the user based solely on sparse rating matrix. A new comprehensive item similarity measurement algorithm based on information entropy is proposed in this paper to dispose the data sparse problem. Meanwhile, taking into account the user＇s interest will change over time, and the change is not same in different user groups, this paper put forward the weight of adapt to different user interest changes, which is inspired by Ebbinghaus＇s memory rule. The performed experiment based on the dataset of movielens shows that the modified algorithm can not only alleviate the problem of rating data sparse, but also can improve the accuracy of the algorithm.