The semantic social network is a type of network that has enormous nodes and intricate semantic information; however, the conventional community detection algorithms were unable to give the expected coherent communities in its place. We present a clustering community detection algorithm that is based on the Hybrid Honey Badger Optimization Modified Latent Dirichlet Algorithm (HBOmLDA) model in order to solve the problem of detecting semantic social networks. We use the Particle Swam Optimization (PSO), which is able to make quantitative parameters map from semantic information to semantic space. Given that the semantic model is an LDA model, this is necessary. Then, in order to solve the problem of overlapping community detection, we present a HBO strategy that incorporates a semantic relation. In the end, we establish semantic modularity, in order to evaluate the semantic communities that were found. The experimental analysis demonstrates that the Hybrid HBOmLDA model’s validity and feasibility, in addition to the semantic modularity, the search space is also optimized using Meta-Heuristics Algorithm..

LDA, HBO, Community Detection, Meta-Heuristics Algorithm.