Hybrid collaborative filtering for high-involvement products: A solution to opinion sparsity and dynamics
详细信息 查看全文
- 作者：Cuiqing Jianga ; jiangcuiq@163.com" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Rui Duana ; duanrui0401@163.com" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Hemant K. Jainb ; jain@uwm.edu" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Shixi Liua ; c ; liusxchuz@163.com" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Kun Lianga ; liangkun_fd@163.com" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae
- 关键词：High-involvement products ; Hybrid collaborative filtering ; Sparsity problem ; Online opinion dynamics
- 刊名：Decision Support Systems
- 出版年：2015
- 出版时间：November 2015
- 年：2015
- 卷：79
- 期：Complete
- 页码：195-208
- 全文大小：1456 K
文摘
- •
A Review-based Hybrid CF (RHCF) is designed to overcome the rating sparsity problem of high-involvement products.
- •
From both ratings and reviews the temporal and sequential dynamics of online opinions are analyzed.
- •
Dynamics of online ratings and reviews are considered in the design of the recommendation method to avoid the possible biases.
- •
The experiment, based on real-world datasets, demonstrates the superior performance of the RHCF method.