Social influence is a widely accepted phenomenon in social networks, and it has been studied by researchers from various perspectives, including social psychology, sociology, marketing, and computer science, just to name a few. During the past decade, the emergence and fast growth of social media sites (such as Facebook and Twitter) have enabled the measurement, quantitative analysis, and modeling of social influence at a large scale. Therefore, it is essential to re-evaluate these developed algorithms and models in the new era of big data. In this chapter, we review research on social influence analysis in the big data era, with a focus on the computational perspective.We first present the statistical measurements of social influence. Then, we introduce the algorithms and models to characterize the propagation of social influence. Next, we present the issues related to the optimization of the propagation of social influence. In addition, we review research on the diffusion of network influence, which is closely related to the studies of the forecasting and influencing/contagion of information. Towards the end of this chapter, we also discuss the envisioned opportunities and challenges.
Introduction
Social influence analysis is an intuitive and well-accepted phenomenon by researchers for decades [1, 2]. Since social influence plays a key role in social life and decision making, as discovered by Katz and Lazarsf in the 1950s [3], theories and models have been developed from various perspectives by researchers in many different areas, including sociology, computer science, and management science, etc. With the popularity of social network services, increasing computer science researchers are paying more attention to this field. Social influence has extensive qualitative and quantitative applications, which have been well studied in sociology and computer science. For example, public opinion leaders affect numerous fans, and their opinions are quickly spread to a large population. Since they play an essential role in information dissemination, many studies focused on the identification of those users [4–6]. Social influence analysis has also been applied to other fields, such as recommendation systems [7], information propagation in social networks [1, 8–11], link prediction [12–14], viral marketing [15–21], public health [22, 23], expert discovery [24, 25], detection of emergent events [26], and advertising [27], just to name a few. In this chapter, we focus on the “social influence analysis” based on social networks such as Twitter, Facebook, and Weibo.