Improved Recommendation Algorithm Research Based on Cloud Model and Time-Weighted
Lixia Han, Gao, Ling, Feifei Yang · ASME Press eBooks · 2011
In order to offer recommendation resources of higher quality and better accuracy, this paper proposes an algorithm based on cloud model and time-weighted, which overcomes the shortcomings of traditional similarity calculation and solves the problem which results from the variability of interest of users with time. Experiments on datasets show that the algorithm has gained higher accuracy than traditional algorithm; at the same time, this algorithm can solve data sparsity and alterability of user's interest with time. So it performs great advancement in quality and accuracy of recommendation service.