Research on User Behavior with Collaborative Recommendation Based on Social Network

Abstract

With the rise of social networks and the existence of a large number of active users, a large number of data recording daily life are constantly generated. This kind of data has the characteristics of interactivity, real-time and sociality, which implies a lot of valuable information. Therefore, social network data has a very positive role in promoting the construction of smart city. In this paper, users' annotation information of the item in social network is used. Similarity model is established with appropriate similarity calculation method. Users' social relationship in social network is used to establish trust relationship and model for users. Personalized recommendation algorithm is improved to dig user behavior data, analyze users' interests and hobbies. Information of interest or item can be recommended to meet the needs of users to the maximum extent. And the stickiness of users to the website is increased. Based on the widely used recommendation algorithm, this paper analyzes the user behavior with trust-based collaborative recommendation algorithm based on user relationship of social network.