Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | https://arxiv.org/abs/2601.22856 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915763553566720 |
|---|---|
| author | Zuo, Yilong Li, Xunkai Zhang, Zhihan Dai, Qiangqiang Li, Ronghua Wang, Guoren |
| author_facet | Zuo, Yilong Li, Xunkai Zhang, Zhihan Dai, Qiangqiang Li, Ronghua Wang, Guoren |
| contents | Multimodal Attributed Graphs (MAGs) have been widely adopted for modeling complex systems by integrating multi-modal information, such as text and images, on nodes. However, we identify a discrepancy between the implicit semantic structure induced by different modality embeddings and the explicit graph structure. For instance, neighbors in the explicit graph structure may be close in one modality but distant in another. Since existing methods typically perform message passing over the fixed explicit graph structure, they inadvertently aggregate dissimilar features, introducing modality-specific noise and impeding effective node representation learning. To address this, we propose OptiMAG, an Unbalanced Optimal Transport-based regularization framework. OptiMAG employs the Fused Gromov-Wasserstein distance to explicitly guide cross-modal structural consistency within local neighborhoods, effectively mitigating structural-semantic conflicts. Moreover, a KL divergence penalty enables adaptive handling of cross-modal inconsistencies. This framework can be seamlessly integrated into existing multimodal graph models, acting as an effective drop-in regularizer. Experiments demonstrate that OptiMAG consistently outperforms baselines across multiple tasks, ranging from graph-centric tasks (e.g., node classification, link prediction) to multimodal-centric generation tasks (e.g., graph2text, graph2image). The source code will be available upon acceptance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_22856 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | OptiMAG: Structure-Semantic Alignment via Unbalanced Optimal Transport Zuo, Yilong Li, Xunkai Zhang, Zhihan Dai, Qiangqiang Li, Ronghua Wang, Guoren Machine Learning Multimodal Attributed Graphs (MAGs) have been widely adopted for modeling complex systems by integrating multi-modal information, such as text and images, on nodes. However, we identify a discrepancy between the implicit semantic structure induced by different modality embeddings and the explicit graph structure. For instance, neighbors in the explicit graph structure may be close in one modality but distant in another. Since existing methods typically perform message passing over the fixed explicit graph structure, they inadvertently aggregate dissimilar features, introducing modality-specific noise and impeding effective node representation learning. To address this, we propose OptiMAG, an Unbalanced Optimal Transport-based regularization framework. OptiMAG employs the Fused Gromov-Wasserstein distance to explicitly guide cross-modal structural consistency within local neighborhoods, effectively mitigating structural-semantic conflicts. Moreover, a KL divergence penalty enables adaptive handling of cross-modal inconsistencies. This framework can be seamlessly integrated into existing multimodal graph models, acting as an effective drop-in regularizer. Experiments demonstrate that OptiMAG consistently outperforms baselines across multiple tasks, ranging from graph-centric tasks (e.g., node classification, link prediction) to multimodal-centric generation tasks (e.g., graph2text, graph2image). The source code will be available upon acceptance. |
| title | OptiMAG: Structure-Semantic Alignment via Unbalanced Optimal Transport |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2601.22856 |