Entropy-Based Social Influence Evaluation in Mobile Social Networks

Abstract

How to evaluate social influence of one user on other users in mobile social networks becomes increasingly important. It can help to identify the influential users in mobile social networks. In addition, it can also provide important insights into the design of social platforms and applications, such as viral marketing, domain expert finding, and advertising. In this paper, we present a framework to quantitatively measure social influence in mobile social networks. In the proposed framework, social influence is a measure of uncertainty with its value represented by entropy. We establish a social relationship graph based on the social network theory that addresses the basic understanding of social influence. Based on the social relationship graph, we present an evaluation model on social influence using information entropy to describe the complexity and uncertainty of social influence. We evaluate the performance of our solution using a customized program based on a real-world SMS/MMS-based communication data set. The numerical simulations and analysis show that the proposed influence evaluation strategies can characterize the social influence of mobile social networks effectively and efficiently.