Identification of Topologies by Using Harmonic Centrality in Huge Social Networks

Abstract

Social network analysis plays vital role in many business applications, they are rich in information. All business applications are looking forward to use this data to increase their business activities. To understand behavior of huge social networks, it is necessary to understand the underlying topology of the network. Network topologies play a major role in analyzing the social networks. Discovering topologies of a graph with fewer nodes is not an issue, but since social networks are large in size with millions of nodes, discovering topologies from such networks is challenging. In this paper we are trying to find the strength of the topologies in a network using different network metrics. We have considered different types of graph topologies like ring, mesh, tree and star of a network. By observing different metrics of similar graphs we have defined relations among different metrics in a structure. Here we are using a metric called harmonic centrality to extend our approach to unconnected graphs. As all real time networks are not fully connected to discover topologies. This approach provides an efficient way of discovering topology of the networks. We also discuss the performance with different data sets and We have analyzed the effect of different types of network topology on other topologies. The results obtained are promising, and can be used in any large and also dynamic graphs.