Dense Subgraph Clustering and a New Cluster Ensemble Method

TL;DR

This paper introduces DSC-Flow-Iter, a dense subgraph community detection algorithm, and a novel ensemble method combining it with modularity-based clustering to enhance accuracy on synthetic networks.

Contribution

The paper presents DSC-Flow-Iter, a new dense subgraph extraction algorithm, and a cluster ensemble technique that improves community detection accuracy by combining different methods.

Findings

DSC-Flow-Iter is competitive with leading methods.

The ensemble method outperforms individual algorithms.

The pipeline improves accuracy on synthetic networks.

Abstract

We propose DSC-Flow-Iter, a new community detection algorithm that is based on iterative extraction of dense subgraphs. Although DSC-Flow-Iter leaves many nodes unclustered, it is competitive with leading methods and has high-precision and low-recall, making it complementary to modularity-based methods that typically have high recall but lower precision. Based on this observation, we introduce a novel cluster ensemble technique that combines DSC-Flow-Iter with modularity-based clustering, to provide improved accuracy. We show that our proposed pipeline, which uses this ensemble technique, outperforms its individual components and improves upon the baseline techniques on a large collection of synthetic networks.