Who is really in my social circle?: Mining social relationships to improve detection of real communities

Abstract

Tie strength allows to classify social relationships and identify different types of them. For instance, social relationships can be classified as persistent and similar based respectively on the regularity with which they occur and the similarity among them. On the other hand, rare and somewhat similar relationships are random and cause noise in a social network, thus hiding the actual structure of the network and preventing an accurate analysis of it. In this article, we propose a method to handle social network data that exploits temporal features to improve the detection of communities by existing algorithms. By removing random relationships, we observe that social networks converge to a topology with more pure social relationships and better quality community structures.