Efficient Algorithms for Social Network Coverage and Reach

TLDR: This paper proposes approximation based approaches to solve two specific problems related to information flow in social network applications and shows that these approaches outperform existing approaches using both a theoretical analysis and experimental results.

Abstract: Social networks, though started as a software tool enabling people to connect with each other, have emerged in recent times as platforms for businesses, individuals and government agencies to conduct a number of activities ranging from marketing to emergency situation management. As a result, a large number of social network analytics tools have been developed for a variety of applications. A snapshot of social networks at any particular time, called a social graph, represents the connectivity of nodes and potentially the flow of information amongst the nodes (or vertices) in the graph. Understanding the flow of information in a social graph plays an important role in social network applications. Two specific problems related to information flow have implications in many social network applications: (a) finding a minimum set of nodes one has to know to recover the whole graph (also known as the vertex cover problem) and (b) determining the minimum set of nodes one required to reach all nodes in the graph within a specific number of hops (we refer this as the vertex reach problem). Finding an optimal solution to these problems is NP-Hard. In this paper, we propose approximation based approaches and show that our approaches outperform existing approaches using both a theoretical analysis and experimental results.