Deep Attention-guided Graph Clustering with Dual Self-supervision

TL;DR

This paper introduces DAGC, a novel deep graph clustering method that employs multi-scale feature fusion and dual self-supervision to significantly improve clustering performance over existing methods.

Contribution

The paper proposes a new deep graph clustering approach with attention-guided multi-scale feature fusion and dual self-supervision, enhancing discriminative feature learning and clustering accuracy.

Findings

Outperforms state-of-the-art methods on six benchmark datasets.

Achieves more than 18.14% improvement in ARI over the best baseline.

Demonstrates the effectiveness of dual self-supervision in deep clustering.

Abstract

Existing deep embedding clustering works only consider the deepest layer to learn a feature embedding and thus fail to well utilize the available discriminative information from cluster assignments, resulting performance limitation. To this end, we propose a novel method, namely deep attention-guided graph clustering with dual self-supervision (DAGC). Specifically, DAGC first utilizes a heterogeneity-wise fusion module to adaptively integrate the features of an auto-encoder and a graph convolutional network in each layer and then uses a scale-wise fusion module to dynamically concatenate the multi-scale features in different layers. Such modules are capable of learning a discriminative feature embedding via an attention-based mechanism. In addition, we design a distribution-wise fusion module that leverages cluster assignments to acquire clustering results directly. To better explore the discriminative information from the cluster assignments, we develop a dual self-supervision solution consisting of a soft self-supervision strategy with a triplet Kullback-Leibler divergence loss and a hard self-supervision strategy with a pseudo supervision loss. Extensive experiments validate that our method consistently outperforms state-of-the-art methods on six benchmark datasets. Especially, our method improves the ARI by more than 18.14% over the best baseline.