Experimental Analysis of Algorithms for Community Detection

Abstract

This is an exciting new approach because by understanding the anatomy of networks, you can get a valuable framework to define related phenomena, from social and technological systems to all sorts of other complex systems found in the real world. Community structure is an essential attribute of complex networks, which has been an active field of research for decades. Understanding the structure of networks is a crucial problem, and community detection algorithms are the most utilized strategies for that. Detecting communities is fundamental for understanding their structure, function, evolution, and dynamics. As a result, the concept of community structure has attracted significant interest in recent years. When thinking about product recommendations, it is crucial to pay attention to finding sub-networks within the co-purchasing network. Keeping this in mind, we made the decision to start research that would involve analyzing four methods for community discovery. We wanted to select some real-world data for our experiment, so we used the Amazon co-purchasing network datasets to test the algorithms we had selected. In addition, we plan to investigate future work that integrates machine learning techniques with algorithms for detecting communities. In this situation, factors such as run time, efficiency, and modularity scores become significant. As a result, our project will concentrate on assessing the run time and modularity ratings of each algorithm in particular.