An Efficient Approach for Community Detection in Complex Social Networks Based on Elephant Swarm Optimization Algorithm

An Efficient Approach for Community Detection in Complex Social Networks Based on Elephant Swarm Optimization Algorithm

Khaled Ahmed (Cairo University, Egypt), Aboul Ella Hassanien (Cairo University, Egypt & Scientific Research Group in Egypt (SRGE), Egypt) and Ehab Ezzat (Cairo University, Egypt)
Copyright: © 2017 |Pages: 14
DOI: 10.4018/978-1-5225-2229-4.ch047
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Complex social networks analysis is an important research trend, which basically based on community detection. Community detection is the process of dividing the complex social network into a dynamic number of clusters based on their edges connectivity. This paper presents an efficient Elephant Swarm Optimization Algorithm for community detection problem (EESO) as an optimization approach. EESO can define dynamically the number of communities within complex social network. Experimental results are proved that EESO can handle the community detection problem and define the structure of complex networks with high accuracy and quality measures of NMI and modularity over four popular benchmarks such as Zachary Karate Club, Bottlenose Dolphin, American college football and Facebook. EESO presents high promised results against eight community detection algorithms such as discrete krill herd algorithm, discrete Bat algorithm, artificial fish swarm algorithm, fast greedy, label propagation, walktrap, Multilevel and InfoMap.
Chapter Preview
Top

Community Detection Problem

The proposed problem is to define dynamically the number of communities within complex social network graph (G), which contains a set of Nodes (N) and edges(E), which illustrated in equation .1, that fulfill the quality function F(S), which illustrated in equation 2 (Ahmed et al.2015). There are a lot of challenges to apply community detection process in social network data such as heterogeneity and evaluation of this complex networks.

(1)
(2) where is an objective quality.

Complete Chapter List

Search this Book:
Reset