Influential Users Detection in Online Social Networks

Abstract

ocial networks are considered one of the main merits of this era. People worldwide use this online platform to build their own social ties. The key feature behind the success of social networks is microblogging. This fea- ture facilitates the interactions between people around the globe. People use social networking platforms to share their ideas, populate their believes and find other people with the same preferences. Social network users tend to interact with each other by sharing, commenting and reacting to disseminated content. These interactions help in the con- tent spread across the network. The dynamics of user interest in the dis- seminated content leads to the clustering of social network users to varying groups (communities) called “interest groups”. The analysis of users be- haviour raises some crucial questions about who is responsible for content spread, the roles played by users in an interest group, the user rank based on his/her role and the rank of the interest group as a whole. Our research objective is to propose ranking models that take into consider- ation the dynamic nature of social networks topology and the users interest to tackle the previously mentioned limitations. In order to achieve this ob- jective, four models are proposed. First, Influence Ranking Model (IRM) which aims to rank all the social network users based on their interactivities.