Reliable User Profile Analytics and Discovery on Social Networks

Abstract

In this paper, we introduce heterogeneous methods to analyze and discover user profiles on Online Social Networks (OSNs).We are the first to investigate such methods to profile users on multiple OSNs (Facebook, Twitter, Google+, etc.). In addition, we perform reliable analytics, i.e., users in the datasets are identical. Deeply speaking, if we have a dataset of n number of user profiles on Facebook, we do not analyze n different profiles on corresponding OSN. However, we first perform a user Profile Matching (PM) task from a seed dataset (Facebook for instance) and then match these profiles inside this dataset to their corresponding profiles on other OSNs, then we start our User Profile Analysis and Discovery task (UPAD). We show that our UPAD methods uncover very interesting facts about OSN users.