Community Detection in Scientific Collaboration Networks using the Louvain Method

This research aims to identify the collaboration of authors in scientific publication networks, which is increasingly important due to the rapid growth of research output using the Louvain method. Understanding the structure of collaboration can help reveal the formation of research communities and the key actors within them. To achieve this, we use three centrality metrics, each of which has a function to help understand collaboration between authors. Authors with ID P107 in degree centrality, authors with ID 177 in betweenness centrality and authors with ID P174 closeness centrality. In applying the Louvain method, the results show a fairly high modularity score of 0.7689, which indicates optimal separation between clusters. This finding shows that although Louvain is quite efficient in separating communities.