Attributed Graph Clustering: A Deep Attentional Embedding Approach
This addresses the challenge of attributed graph clustering for researchers and practitioners in network analysis, offering an incremental improvement by unifying embedding and clustering processes.
The paper tackles the problem of suboptimal performance in graph clustering due to two-step frameworks by proposing a goal-directed deep learning approach called Deep Attentional Embedded Graph Clustering (DAEGC), which integrates graph embedding and clustering into a unified framework and demonstrates superiority over state-of-the-art algorithms in experiments.
Graph clustering is a fundamental task which discovers communities or groups in networks. Recent studies have mostly focused on developing deep learning approaches to learn a compact graph embedding, upon which classic clustering methods like k-means or spectral clustering algorithms are applied. These two-step frameworks are difficult to manipulate and usually lead to suboptimal performance, mainly because the graph embedding is not goal-directed, i.e., designed for the specific clustering task. In this paper, we propose a goal-directed deep learning approach, Deep Attentional Embedded Graph Clustering (DAEGC for short). Our method focuses on attributed graphs to sufficiently explore the two sides of information in graphs. By employing an attention network to capture the importance of the neighboring nodes to a target node, our DAEGC algorithm encodes the topological structure and node content in a graph to a compact representation, on which an inner product decoder is trained to reconstruct the graph structure. Furthermore, soft labels from the graph embedding itself are generated to supervise a self-training graph clustering process, which iteratively refines the clustering results. The self-training process is jointly learned and optimized with the graph embedding in a unified framework, to mutually benefit both components. Experimental results compared with state-of-the-art algorithms demonstrate the superiority of our method.