The power of network based framework for tackling various independent application domains

Abstract

Recently, with the advent of popular social networking platforms, a completely new area of research called Social Network Analysis (SNA) has emerged. SNA is dedi­cated to the study of techniques for storing, manipulating and analyzing very large graphs of up to billions of vertices and edges. In SNA terminology the graphs are commonly referred to as social networks since they model the relationships between individuals. In this terminology a social network consists of a set of actors corre­sponding to individuals connected to each other by links representing the relation­ships. Community detection is one of the major topics that fall within the scope of SNA which aims at unfolding communities of actors who are densely connected to each other than to actors outside the community. Many powerful algorithms have been proposed to address the community detection problem in large networks. While different applications of the graph-based algorithms have been well studied, a little work has been done on the application of the novel SNA algorithms on the problems in other fields of research. In this article, I examine the application of graph-based and SNA algorithms in a wide variety of domains, namely, database design and data distribution, multidimensional data allocation, outlier detection, and social media data summarization. My experiments prove that when combined with commonly used data mining techniques the algorithms can provide powerful solutions to the problems.