An approach to mining information from telephone graph using graph mining techniques

Among various properties of social network, one of the important properties is to study strong community effect where social entity in a network forms a group which is closely connected. Groups formed out of such properties are communities, clusters, cohesive subgroups or modules. The authors have observed that individuals interact more frequently within a group rather than group interaction. Detection of similar groups in a social network is known as community detection. Finding such type of communities and analyzing, helps in knowledge and pattern mining. This paper focuses on methods to study a real world social network communications using the basic concepts of graph theory. For this purpose, the authors have considered telephone network. The authors have proposed an algorithm for extracting different network provider's sub-graphs, weak and strong connected sub-graphs and extracting incoming and outgoing calls of subscribers which have direct application for studying the human behavior in telephone network. The proposed algorithm has been implemented in C++ programming language and obtained satisfactory result.