Fusion Based Approach to Discovering Social Circles in Ego Networks

Abstract

Communication among people via social network generates enormous amount of data every day. This data can be valuable if they are properly organized. Only with the proper organization of these data into scalable structures, we will be able to find pattern in these data. A very important step in organizing these data is to organize the social network itself. The main challenge for users of social media is to organize their networks and content generated by them by categorizing their friends and the people they follow into different social circles. Social circles are actually lists of people from a person’s friends that can be used for content filtering, privacy and sharing information only among those people in his/her circle. In this paper, we propose a method to automatically discovery of a person’s potential overlapping social circles by usage of different community detection algorithms and clustering methods. Our approach performed significantly well as compared to classical community detection methods.

Publication
In International Conference on Machine Learning and Cybernetics, IEEE

Related