Emmanuel Flores Huicochea, National Polytechnic Institute, Mexico

In mobile social networks, the process of detecting community structures is crucial for untangling the intricate web of connections found within large datasets. This is not only important for managing these extensive networks but also for gaining a deeper understanding of their various components. One effective method for achieving this is Link Community Detection (LCD), which identifies community structures by analyzing the similarities between neighboring edges and assessing partition density. This powerful technique can uncover communities that contain overlapping nodes, enhancing the granularity of the analysis, and is well-suited for dynamic networks that evolve over time. However, when a network undergoes significant changes in a short period, the ability to accurately reveal the identities of these communities can be compromised. To address this challenge, we propose an innovative identification algorithm specifically designed for evolutionary communities. The essence of our approach lies in it tracking communities throughout the detection process, comparing the similarities between established communities and new elements as they emerge. This method not only improves the accuracy of community representation but also facilitates a richer understanding of the dynamic interactions within the network, enabling more informed management and analysis strategies.

APA:
Huicochea, E. F. (2021). An Effective Algorithm for Identifying Evolving Communities in Mobile Social Networks. Journal of Science and Engineering Management, 2(2), 11-22. https://doi.org/10.33832/jsem.2021.2.2.02