Social networks have emerged as one of the most popular platforms for the analysis of communication. The detection of communities in these networks play a crucial role in today’s world. The Cosine Shared Link Method (CSLM) is a greedy technique to detect communities in the network. It detects partition in large networks using Cosine similarity measure. This method follows a procedure inspired by the state-of-the-art Louvain method, effectively maximizing modularity. In this paper, the Cosine Shared Link Method was studied and its performance was compared against other existing community detection algorithms. The evaluation was done using modularity and the four algorithms were implemented on five social networks in an attempt to analyze the performance of CSLM.