Balanced Multi-Relational Graph Clustering

Zhixiang Shen, Haolan He, zhao kang

Abstract: Multi-relational graph clustering has demonstrated remarkable success in uncovering underlying patterns in complex networks. Representative methods manage to align different views motivated by advances in contrastive learning. Our empirical study finds the pervasive presence of imbalance in real-world graphs, which is in principle contradictory to the motivation of alignment. In this paper, we first propose a novel metric, the Aggregation Class Distance, to empirically quantify structural disparities among different graphs. To address the challenge of view imbalance, we propose Balanced Multi-Relational Graph Clustering (BMGC), comprising unsupervised dominant view mining and dual signals guided representation learning. It dynamically mines the dominant view throughout the training process, synergistically improving clustering performance with representation learning. Theoretical analysis ensures the effectiveness of dominant view mining. Extensive experiments and in-depth analysis on real-world and synthetic datasets showcase that BMGC achieves state-of-the-art performance, underscoring its superiority in addressing the view imbalance inherent in multi-relational graphs. The source code and datasets are available at https://github.com/zxlearningdeep/BMGC.

Primary Subject Area: [Content] Multimodal Fusion

Secondary Subject Area: [Experience] Multimedia Applications

Relevance To Conference: This work contributes to multimedia/multimodal processing by addressing the critical issue of view imbalance in complex data, particularly relevant in multimedia/multi-relational data where various modalities are represented as graphs. By considering multiple relations or views simultaneously, we harness complementary information from different modalities, significantly advancing multimedia processing with a balanced clustering solution. This balance is crucial in multimedia scenarios where certain views may dominate due to inherent characteristics. Our approach enhances understanding and utilization of multi-relational data, aligning with the ACM MM conference's focus on promoting research in multimedia and multimodal processing. This research underscores the significance of maintaining equilibrium in representing various views for comprehensive and unbiased interpretation of multimedia and multimodal processing where different modalities interact.

Submission Number: 3397