A survey of user classification in social networks

Social networks is a virtual communities and networking platform, where people can create and share views, ideas and experiences freely. Social networks like the real world, filled with colorful crowd. How to use user behavior attributes, basic information, and interactive content in social media to classify users has become a hot topic in the field of social network analysis. In this paper, we present a detailed review in various types of user classification in social networks. We present most common methods based on machine learning and non-machine learning in user classification. The algorithms based on machine learning includes: Bayesian, Decision Tree, Logistics, SVM and KNN. The method of non-machine learning includes: concept of entropy and based on user similarity.