Network community detection through enhanced effective distance information: A scalable and parameter-free mechanism

Network community detection plays a pivotal role in understanding and unveiling the inherent structure and concealed information within complex networks. These networks encompass various systems such as online social networks, transport systems, and biological networks. Given the complexity and extensive size of these networks, detecting communities within them poses significant challenges. Numerous community detection methods have been proposed. These methods can be broadly categorized into local and global community detection algorithms. Local detection algorithms rely on localized information within a specific section of the network, while global algorithms use information from across the entire network. However, these methods often necessitate preliminary parameter adjustments, which not only complicates the detection process but also increases the computational overhead. In light of these challenges, there is a critical need for a reliable, computationally efficient approach to community detection. To address this, the paper proposes an integrated framework called the effective distance information-enhanced mechanism (EDIM). EDIM methodology combines the advantages of both local and global information, thereby eliminating the requirement for initial parameter adjustments. EDIM framework operates in three stages. The first stage involves analyzing the local and path information of nodes to establish their significance within the network. This is followed by the creation of node pairs to identify their second-order neighbors in the second stage. The final stage comprises the actual detection and identification of communities. Our simulations and experimental results exhibit that EDIM framework can accurately detect high-quality communities in both synthetic and real-world networks.