A Survey of Graph-based Clustering Techniques

Abstract

Graph clustering is a powerful technique used to identify and group similar nodes within a complex network structure. This procedure involves segmenting the graph into distinct groups, with the nodes in each group having strong interconnections or similar characteristics. In this survey, we will highlight a variety of existing approaches of graph clustering, including spectral clustering, modularity optimization and hierarchical clustering, to efficiently discover meaningful clusters, facilitating analysis and decision-making in diverse fields ranging from data science and network analysis to social sciences and bioinformatics.