Modeling and Comparing the Influence of Neighbors on the Behavior of Users in Social and Similarity Networks

摘要: Social networks are becoming more and popular with the advent of numerous online social networking services. In this paper, we explore rating networks, which record not only relations but also user ratings for items. We analyze model effects influence correlational in such based on coefficients that measure degree a network. distinguish two types behavior: adopting an item value item. propose models to neighbors both adoption behavior users. Our experiments demonstrate has much stronger impact than influence. global entire However, there local differences, i.e. certain users have others. To effect, introduce novel concept authority individual objective way evaluate by injecting it into simple recommender system.