Abstract:Benefiting from the intrinsic supervision information exploitation capability, contrastive learning has achieved promising performance in the field of deep graph clustering recently. However, we observe that two drawbacks of the positive and negative sample construction mechanisms limit the performance of existing algorithms from further improvement. 1) The quality of positive samples heavily depends on the carefully designed data augmentations, while inappropriate data augmentations would easily lead to the semantic drift and indiscriminative positive samples. 2) The constructed negative samples are not reliable for ignoring important clustering information. To solve these problems, we propose a Cluster-guided Contrastive deep Graph Clustering network (CCGC) by mining the intrinsic supervision information in the high-confidence clustering results. Specifically, instead of conducting complex node or edge perturbation, we construct two views of the graph by designing special Siamese encoders whose weights are not shared between the sibling sub-networks. Then, guided by the high-confidence clustering information, we carefully select and construct the positive samples from the same high-confidence cluster in two views. Moreover, to construct semantic meaningful negative sample pairs, we regard the centers of different high-confidence clusters as negative samples, thus improving the discriminative capability and reliability of the constructed sample pairs. Lastly, we design an objective function to pull close the samples from the same cluster while pushing away those from other clusters by maximizing and minimizing the cross-view cosine similarity between positive and negative samples. Extensive experimental results on six datasets demonstrate the effectiveness of CCGC compared with the existing state-of-the-art algorithms.

Effective Sample Pairs Based Contrastive Learning for Clustering

Doubly Contrastive Deep Clustering

Deep Clustering by Graph Attention Contrastive Learning

A self-supervised learning of semantic feature consistency for image clustering

Strongly Augmented Contrastive Clustering

Unsupervised node clustering via contrastive hard sampling

Deep Image Clustering by Fusing Contrastive Learning and Neighbor Relation Mining.

Combining core points and cluster-level semantic similarity for self-supervised clustering

Deep Clustering with Hybrid-Grained Contrastive and Discriminative Learning

Cluster-guided Contrastive Graph Clustering Network

Multiple Contrast Clustering Network

Deep Robust Clustering by Contrastive Learning

Deep Clustering by Semantic Contrastive Learning

Contrastive Deep Embedded Clustering

Safe Contrastive Clustering

Deep Contrastive Clustering via Hard positive sample Debiased

Exploring Non-Contrastive Representation Learning for Deep Clustering

A Debiased Graph Clustering Approach Using Dual Contrastive Learning

Multi-CC: A New Baseline for Faster and Better Deep Clustering

Deep Temporal Contrastive Clustering