A Novel Dynamic Community Detection Model in Social Networks

Social networks are becoming more and more popular and important. A key feature is that they are dynamic and develop with time. As a result, their community evolves along with the social networks. Detecting community structure dynamically is an important research topic. Specifically, how social networks along with communities evolve is still not revealed. In this paper, we propose the importance of individual changing topics that is a key factor in community evolution. this paper designs a novel dynamic community detection model of social networks by considering the natural evolutionary features DCDIT (Dynamic Community Detection Motivated by Individual Changing Topics). The key mechanism is that individual changing topics drive network and community structure evolve. Our approach begins by dividing social networks into distinct snapshots. Following this, it identifies community structures through the analysis of varying individual topics and information from the preceding snapshot. Subsequently, it determines the topic distributions that interest users and those prevalent in communities. This model is benchmarked against four advanced baseline models using two datasets. Experimental outcomes reveal that DCDIT surpasses all baseline models in performance.