【摘要】 为提高推荐系统的推荐质量,减小用户评分数据稀疏性对推荐效果的不良影响,提出了一种结合加权信息熵与兴趣度的协同过滤推荐算法。此算法全面考虑公共评分项目数、评分数值差异与数值变化趋势三个方面的因素,结合加权信息熵与用户兴趣度,使度量用户间相似度变得更加准确。仿真实验结果表明:所提算法比基于Pearson相关性系数以及余弦相似性的推荐算法具有更小的平均绝对误差,表明了其可行性和有效性。更多还原

【Abstract】 In order to improve the quality of the recommendation and reduce the negative impacts of sparse user rating data,a new collaborative algorithm is put forward based on weighted information entropy and the rating of user interests.The algorithm considers the number of common rating items,numerical differences and numerical change trend three factors,makes user similarity calculations to be more credible by combining the weighted information entropy with the rating of user interests.The simulation experiment results show that the proposed algorithm is better than the recommendation algorithm based on Pearson correlation and cosine similarity because of smaller Mean Absolute Error(MAE).In conclusion,the new algorithm is feasible and effective.更多还原