On Comparing and Enhancing Common Approaches to Network Community Detection

TL;DR

This paper compares four popular network community detection algorithms, analyzes their mechanics, and proposes enhancements that improve their efficiency and clustering behavior based on evaluations on standard datasets.

Contribution

It provides a detailed comparison of common community detection algorithms and introduces novel enhancements to improve their performance and clustering outcomes.

Findings

Enhanced Louvain Method with deterministic faster version

Improvements to Fastgreedy algorithm implementation

Promising results from self-neighboring in Neighbor Matrix construction

Abstract

In this work, we explore four common algorithms for community detection in networks, namely Agglomerative Hierarchical Clustering, Divisive Hierarchical Clustering (Girvan-Newman), Fastgreedy and the Louvain Method. We investigate their mechanics and compare their differences in terms of implementation and results of the clustering behavior on a standard dataset. We further propose some enhancements to these algorithms that show promising results in our evaluations, such as self-neighboring for Neighbor Matrix constructions, a deterministic slightly faster version of the Louvain Method that favors less bigger clusters and various implementation changes to the Fastgreedy algorithm.