Enhancing Semi-Supervised Multi-View Graph Convolutional Networks via Supervised Contrastive Learning and Self-Training

2025-12-15

AI Summary

To address insufficient cross-view complementary information exploitation, weak feature representation, and poor generalization in semi-supervised learning on multi-view graph data, this paper proposes a novel multi-view Graph Convolutional Network (GCN) framework. Methodologically, it introduces: (1) a synergistic optimization mechanism integrating supervised contrastive loss with pseudo-label-based self-training, enabling end-to-end joint optimization of cross-entropy and contrastive objectives; (2) a dual-graph construction strategy combining k-nearest neighbor (KNN) graphs and semi-supervisedly constructed graphs to enhance topological robustness; and (3) multi-view semantic alignment-driven contrastive learning to explicitly model inter-view complementarity. Extensive experiments on multiple benchmark datasets demonstrate consistent and significant improvements over state-of-the-art methods, with average classification accuracy gains of 2.1–4.7%. The source code is publicly available.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Computer Vision: Multi-modal Vision
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

The advent of graph convolutional network (GCN)-based multi-view learning provides a powerful framework for integrating structural information from heterogeneous views, enabling effective modeling of complex multi-view data. However, existing methods often fail to fully exploit the complementary information across views, leading to suboptimal feature representations and limited performance. To address this, we propose MV-SupGCN, a semi-supervised GCN model that integrates several complementary components with clear motivations and mutual reinforcement. First, to better capture discriminative features and improve model generalization, we design a joint loss function that combines Cross-Entropy loss with Supervised Contrastive loss, encouraging the model to simultaneously minimize intra-class variance and maximize inter-class separability in the latent space. Second, recognizing the instability and incompleteness of single graph construction methods, we combine both KNN-based and semi-supervised graph construction approaches on each view, thereby enhancing the robustness of the data structure representation and reducing generalization error. Third, to effectively utilize abundant unlabeled data and enhance semantic alignment across multiple views, we propose a unified framework that integrates contrastive learning in order to enforce consistency among multi-view embeddings and capture meaningful inter-view relationships, together with pseudo-labeling, which provides additional supervision applied to both the cross-entropy and contrastive loss functions to enhance model generalization. Extensive experiments demonstrate that MV-SupGCN consistently surpasses state-of-the-art methods across multiple benchmarks, validating the effectiveness of our integrated approach. The source code is available at https://github.com/HuaiyuanXiao/MVSupGCN

Problem

• Enhancing multi-view graph convolutional networks with supervised contrastive learning and self-training
• Improving feature representation by combining cross-entropy and contrastive loss functions
• Integrating multiple graph construction methods and pseudo-labeling for better generalization

Innovation

• Combines cross-entropy and supervised contrastive loss for better feature separation.
• Uses KNN and semi-supervised methods to build robust multi-view graphs.
• Integrates contrastive learning with pseudo-labeling to align multi-view embeddings.