基于隐性用户评分的推荐系统
首发时间:2012-11-21
摘要:目前电子商务推荐系统在理论和实践中都得到了很大发展,但现有的推荐系统普遍存在着用户评分稀疏性问题以及用户内在评分标准偏差问题,难以有效建立用户兴趣模型。本文提出了一种基于隐性用户评分的自适应推荐算法,阐述了推荐系统的整体架构,包括个性化用户数据的收集、用户评分标准化、满意度计算与反馈过程。实验数据表明,该推荐算法对用户兴趣能做出较准确的判断,同时能及时有效地掌握用户兴趣偏移,从而改善用户体验,增加用户粘性,提升推荐的精确度。
关键词: 计算机应用技术 推荐系统 隐性评分 兴趣模型 协同过滤 自适应
For information in English, please click here
Implicit user rating-based recommender system
Abstract:E-commerce recommendation system in theory and practice have been great development, but existing recommendation system common user rating sparsity problem, as well as the internal user ratings standard deviation, it is difficult to effectively build user interest model. This paper presents an adaptive recommendation algorithm based on the recessive user ratings on the overall structure of the recommendation system, including personalized user data collection, user ratings, standardization, calculation of satisfaction and feedback process. The experimental data show that the recommendation algorithm on user interest can make more accurate judgments, while master user interest in a timely and effective manner to offset, to improve the user experience, increase user stickiness, improve the accuracy of the recommended.
Keywords: Computer application technology Recommended system implicit rating interest model collaborative filtering adaptive
论文图表:
引用
No.****
同行评议
共计0人参与
勘误表
基于隐性用户评分的推荐系统
评论
全部评论0/1000