Pure expansion-based local community detection

The identification of overlapping communities in complex networks is crucial for network analysis. However, it becomes extremely challenging when dealing with large-scale networks, like social media platforms or news networks, with billions of entities. This task involves identifying patterns within the vast web of connections among vertices and edges, making it complex and demanding. Despite the development of various community detection methods, the problem persists without any significant improvement. Some algorithms may generate communities with low quality, while others may have resolution issues or scalability challenges, especially when analysing large-scale networks. To overcome these research challenges, we require a method that is highly stable, has low time complexity, and efficiently identifies overlapping communities in networks. Therefore, we introduce the pure expansion-based local community detection algorithm (PE-LCD), designed to identify overlapping communities within complex networks through the selection and expansion of seeds, all without the need for pre-existing community knowledge. Initially, we select a seed node based on its centrality within the network, and subsequently expand this seed node into a community through the node and edge interaction coefficient. To assess the efficacy of PE-LCD, we conducted thorough experiments on both real-world and artificial networks. The results indicate that our proposed approach surpassed state-of-the-art techniques, successfully identifying overlapping communities within complex networks.