Research On Community Detection Algorithms Based On Intelligent Computation In Complex Networks

Community detection is one of the most challenging problems in complex network analysis.This problem attracts an amount of interest from various scientific fields such as biology,social network,and physics.In the past few decades,numerous heuristics and exact algorithms have been designed to address this challenging problem.However,most of them are not suitable for large networks,since they require considerable computing time.Contrary to the recent trend in the community detection literature,where complex nature-inspired methods are often proposed,in this paper,we present a simple metaheuristic approach based on the iterated local search(ILS)algorithm which have been applied with great success to the related problems.Extensive comparative evaluations are carried out against the state-of-the-art techniques for the problem in the literature.The computational results show that ILS can detect communities with high quality and stability.Structural balance,one of the most significant properties in ensembles of signed networks reflects the potential tensions and conflicts of signed networks.The solution of clustering problem is an important criterion to measure the level of balance in signed networks.The continuous growth of signed networks makes exact methods not suitable for being used,since they require large computing time.In this paper,we propose a metaheuristic approach based on the iterated greedy(IG)algorithm for partitioning signed networks.Experimental results on the well-known benchmark instances and comparisons with other state-of-art metaheuristics show the efficiency and effectiveness of the proposed algorithm.