Growth Dynamics of 'Socl': an Interest-Based Social Network

Abstract

Online Social Networks have been attracting millions of users who have adopted these sites in their daily lives activities. As the amount of information and the real-time data uploaded to these systems increasing, it becomes a rich source for researchers from industry and academic institutions to study the human behavior patterns for different purposes. The structure of these expanding networks is expected to mirror the structure of real-life communications and relationships in society. In this thesis, we adopted graph theory models and some statistical properties and measurements of the social networks to analyze the dynamic growth in ‘Socl' which is an interest-based social network developed by Microsoft FUSE Labs. We studied the changes in the properties of the users' friendship network quantified based on their interactions and we presented a series of measurements of the users' friendship and interests behavior in the range of time between January and November 2012. We observed interesting evolutionary patterns on the topological structure of the ‘Socl' network as it went from a private project within Microsoft Corporation to a shared project with students from three universities in an invitation-only stage before it released to the public in May 2012. We demonstrated some social network features like scale-free and small world effect. Moreover, we investigated the presence of the Homophily principle among the users, and results indicated that users tend to associate and bond more with others who have similar interests as their own. The findings of these analysis are essential to the understanding of the dynamic change of the structural characteristics of an interest-based social network.

Date of Award	2014
Original language	American English
Supervisor	Iyad Rahwan (Supervisor)