Kernel Searching Strategy for Recommender Searching Mechanism

摘要

A trust-aware recommender system (TARS) is widely used in social media to find useful information. Recommender searching mechanism is an important research issue in TARS. We propose a new searching strategy for recommender searching mechanism of TARS, which named kernel searching strategy. A kernel, which consists of hub nodes of the trust network, is involved in trust propagations. The kernel can be obtained from node degree or node betweenness, take these hub nodes as active users and then finds the recommenders via trust propagations from the kernel, most of the nodes in the network will be covered. Comparing the results of these two methods, the coverage rate of these hub nodes which is obtained from the node degree is almost less than that obtained from the node betweenness. To get better coverage rate, we take both degree and betweenness into consideration. The results show that the combination can get better coverage rate only compared with the node degree. However, the combination has better convergence effect compared with the node betweenness.