Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
5102765 | Physica A: Statistical Mechanics and its Applications | 2017 | 7 Pages |
Abstract
Recommender systems play an important role to help us to find useful information. They are widely used by most e-commerce web sites to push the potential items to individual user according to purchase history. Network-based recommendation algorithms are popular and effective in recommendation, which use two types of elements to represent users and items respectively. In this paper, based on consistence-based inference (CBI) algorithm, we propose a novel network-based algorithm, in which users and items are recognized with no difference. The proposed algorithm also uses information diffusion to find the relationship between users and items. Different from traditional network-based recommendation algorithms, information diffusion initializes from users and items, respectively. Experiments show that the proposed algorithm is effective compared with traditional network-based recommendation algorithms.
Keywords
Related Topics
Physical Sciences and Engineering
Mathematics
Mathematical Physics
Authors
Wenping Ma, Chen Ren, Yue Wu, Shanfeng Wang, Xiang Feng,