A model for reputation rank in online social networks and its applications

Abstract

The volume of information users upload through online social networks (OSNs) is continuously growing. Our focus in this research is in evaluating models to quantify the volume and strengths of interactions between users in OSNs. In our first model, we proposed a reputation rank in OSNs based on a tree graph in which users represent the tree nodes and edges represent their friends' connections or their generated activities. In each OSN such as Facebook, Twitter, LinkedIn, etc. each user will be given a single value that represents the user created activities and friends' interactions with those activities. The model focuses on volumes and statistics of interactions, rather than the content. We also extended the use of cliques' models in OSNs to be normalised or weighted based on the volumes of interactions among clique members. We showed that this can show deeper knowledge of clique relations when comparing it with the classical non-weighted clique models.