Integrating interactions between target users and opinion leaders for better recommendations: An opinion dynamics approach

The social recommender system can accurately recommend information to users, according to their interests based on the characteristics of their social network, however, the interaction between users has not been fully captured in the existing social recommender systems. This study contributes to the literature by proposing a social recommendation method on the basis of opinion dynamics, which captures the information on the interactions between target users and opinion leaders. In our model, the impact of opinion leaders and the evolutionary opinion dynamics between opinion leaders and the target user are integrated to make a recommendation. Experiments based on two real rating datasets, Epinions and FilmTrust were conducted to test the proposed model. The results show that our proposed method can effectively solve the cold-start problem and outperforms the baseline models.