Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
10481537 | Physica A: Statistical Mechanics and its Applications | 2012 | 9 Pages |
Abstract
⺠The patterns of online users' behavior and the effect on recommendation are studied. ⺠We find that users' taste preferences are heterogeneous. ⺠Users' interests for niche items are shown to be highly clustered. ⺠The reshuffling process does not affect the recommendation accuracy for hot items. ⺠The accuracy for niche items drops significantly if we remove users' real patterns.
Related Topics
Physical Sciences and Engineering
Mathematics
Mathematical Physics
Authors
Cheng-Jun Zhang, An Zeng,