.png){kind=icon}
返回
Improving collaborative filtering recommendations by estimating user preferences from clickstream data
.png){kind=icon}
PRE
AI
DOI：10.1016/j.elerap.2019.100877.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
For practical applications of collaborative filtering, we need a user-item rating matrix that encodes user preferences for items. However, estimation of user preferences is inevitably affected by some degree of noise, which can markedly degrade the recommender performance. The primary aim of this research is to obtain a high-quality rating matrix by the effective use of clickstream data, which are a record of a user's page view (PV) history on an e-commerce site. To this end, we use the shape-restricted optimization model for estimating item-choice probabilities from the recency and frequency of each user's previous PVs. Experimental results based on real-world clickstream data demonstrate that higher recommender performance is achieved with our method than with baseline methods for constructing a rating matrix. Moreover, high recommender performance is maintained by our shape-restricted estimation even when only a limited number of training samples are available.
Keyword:
Collaborative filtering
User preference
Rating matrix
Clickstream data
E-commerce
Recommender system
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
6.3
论文数:
2.4K
被引数:
5.9K
机构
引用论文
.png){kind=icon}
Collaborative filtering using orthogonal nonnegative matrix tri-factorization基于正交非负矩阵三分解的协同过滤算法
PREAI
分享收藏Application of Web usage mining and product taxonomy to collaborative recommendations in e-commerce
PREAI
分享收藏
分享收藏Toward the next generation of recommender systems: A survey of the state-of-the-art and possible extensions走向下一代推荐系统: 对最新技术和可能扩展的调查
PREAI
分享收藏A clustering based approach to improving the efficiency of collaborative filtering recommendation
PREAI
分享收藏分享收藏
查看更多内容