Identifying Core Users Based on Trust Relationships and Interest Similarity in Recommender System
Gaofeng Cao, Li Kuang · 2016
With the rapid development of Internet, the explosive growth of information challenges people's capability on finding out items fitting to their own interests. The emergence of recommender system helps users to make decisions to a certain degree. So far, most of the studies pay much attention to designing or improving recommendation algorithms. However, few works consider the extraction of core users with whom recommender systems can generate satisfactory recommendation. In this paper, we propose new approaches to identifying core users based on trust relationships and interest similarity. The trust degree and interest similarity between all user pairs are calculated and sorted first, and two strategies based on frequency and weight of location are used to select core users. Experiments show the effectiveness of the extraction of core users and prove that 20% of core users enable recommender systems to achieve more than 90% of the accuracy of the top-N recommendation.