摘要
传统的基于项目的协同过滤算法离线计算项目相似性,提高了在线推荐速度,但该算法仍然不能解决数据稀疏性所带来的问题,计算出的项目相似性准确度较差,影响了推荐质量。针对这一问题,提出了一种结合类别偏好信息的协同过滤算法。定义了类别偏好相似性,采用类别偏好相似性方法为目标项目找出一组类别偏好相似的候选邻居,在候选邻居中搜寻最近邻,排除了与目标项目共同评分较少项目的干扰,从整体上提高了最近邻搜寻的准确性。在Movie Lens数据集上进行了实验,实验结果表明,新算法的推荐质量较传统的基于项目的协同过滤算法有显著提高。
The traditional item-based collaborative filtering（ CF） algorithm computes item-item similarity offline,so it enhances the real-time performance of recommender system. However,item-based CF algorithm still suffers from the data sparsity problem,as a result that the recommendation quality is poor. To address this issue,this paper proposed a novel CF algorithm combined with class preference information. The proposed algorithm first found out candidate neighbors who were similar to the target item in class preference. Then it searched for nearest neighbors in the candidate neighbor set,which eliminated the interference of the items those had few co-ratings with the target item. Experimental results based on Movie Lens dataset show that the recommendation quality of the new algorithm is significantly improved compared with traditional item-based CF algorithm.
出处
《计算机应用研究》
CSCD
北大核心
2016年第3期669-672,共4页
Application Research of Computers
基金
国家自然科学基金资助项目(71201145)
上海市教育委员会科研创新资助项目(15ZS064)
上海电力学院科研基金资助项目(K2014-037)
上海高校青年教师培养资助计划(zzsdl15115)
关键词
推荐系统
协同过滤
类别偏好
相似性
recommender system
collaborative filtering
class preference
similarity
