摘要
针对家庭用户的电视节目个性化推荐问题,提出一种基于马尔可夫聚类和混合协同过滤(MCL-HCF)算法的混合推荐方法。采用马尔可夫聚类对各个时间段的电视用户进行聚类,产生不同的群组,最小化每个群组里的个体成员和群组整体的偏好差异,再以群组为单位进行电视节目推荐;使用基于物品的协同过滤和基于用户的协同过滤算法分别产生推荐列表;采用基于加权融合的混合推荐算法对两个推荐列表进行处理,得到最终的混合推荐结果。在公开数据集上的实验结果表明,该算法在平衡推荐惊喜度和相关性的同时能够获得令人满意的推荐准确率。
A hybrid recommendation method based on Markov cluster and hybrid collaborative filtering(MCL-HCF)algorithm is proposed to solve the personalized recommendation problem of TV programs for home users.Markov cluster algorithm was used to cluster users in different time slots to generate several groups,which minimized the preference difference between individual members of each group and the entire group.The TV recommendation was applied on these groups.Then,the item-based collaborative filtering and user-based collaborative filtering were used to generate the recommendation list respectively.Finally,a hybrid recommendation algorithm based on weighted fusion was utilized to process the two recommendation lists to obtain the final recommendation results.Experimental results on the open dataset show that the proposed algorithm can achieve satisfactory accuracy while retaining the balance of surprise degree and correlation of recommendation results.
作者
赵宇
刘凤
舒巧媛
韦鹏程
Zhao Yu;Liu Feng;Shu Qiaoyuan;Wei Pengcheng(School of Mathematics and Information Engineering,Chongqing University of Education,Chongqing 400065,China)
出处
《计算机应用与软件》
北大核心
2020年第2期218-225,共8页
Computer Applications and Software
基金
国家自然科学基金项目(11771067)
重庆市教育委员会科学技术研究计划重点项目(KJZDK201801601)
重庆市教委科学技术研究项目(KJQN201801610,KJQN201801605)
重庆市教育科学规划课题(2018-GX-018)。
关键词
个性化推荐
混合推荐
马尔可夫聚类
协同过滤
加权融合
Personalized recommendation
Hybrid recommendation
Markov cluster
Collaborative filtering
Weighted fusion
