Social Network Analysis: Beyond the Greediness in Community Detection Methods

In the context of social network analysis, one of the main challenges concerns community detection. It has attracted the interest and significant attention towards contemporary multidisciplinary research in order to understand the complex structure of networks, unraveling clusters or communities within various interconnected systems. The Louvain algorithm has gained widespread popularity due to its remarkable calculation speed and efficiency, making it one of the most used methods. Although it is celebrated for its efficiency, its greedy nature poses certain challenges, especially in scenarios involving sparsely connected communities. Our study proposes an innovative method that enhances the performance of the Louvain algorithm without increasing its computational complexity. We explore the core applications of this method, seeking to refine and optimize the Louvain algorithm’s performance. By addressing its weaknesses, particularly in handling low-connected communities, our research contributes to developing an improved version of this widely used algorithm. This enhancement is poised to broaden the algorithm’s applicability and reliability, further cementing its status in the field of community detection within complex networks.