Fast and Accurate Community Detection in Social Media Networks with a Multi-Algorithmic Approach

Community detection is an essential technique for uncovering structural patterns in complex networks, such as social, biological, and information networks. In this paper to explore a hybrid approach to community detection by combining the label propagation algorithm and leiden algorithm. Although LPA is well known for its computational effectiveness in large- scale networks, because of its reliance on local label propagation, it frequently results in unstable and fractured communities. High- quality community structures are produced by the Leiden algorithm, which is more computationally expensive than the Louvain method but ensures well-connected communities by optimising modularity. Our goal in combining these two approaches is to take use of LPA's speed for quick community discovery and the Leiden algorithm's refinement powers for increased accuracy and stability. Large or dynamic networks where both real-time speed and community quality are crucial are best suited for the suggested hybrid method. The combination of these algorithms yields more cohesive and modular communities than either one working alone, as evidenced by experimental results on synthetic social network data. The method is a useful tool for large- scale network analysis because it strikes a balance between the necessity for strong and significant community structures and computational efficiency.