Multi-Label Node Classification with Label Influence Propagation
About
Graphs are a complex and versatile data structure used across various domains, with possibly multi-label nodes playing a particularly crucial role. Examples include proteins in PPI networks with multiple functions and users in social or e-commerce networks exhibiting diverse interests. Tackling multi-label node classification (MLNC) on graphs has led to the development of various approaches. Some methods leverage graph neural networks (GNNs) to exploit label co-occurrence correlations, while others incorporate label embeddings to capture label proximity. However, these approaches fail to account for the intricate influences between labels in non-Euclidean graph data. To address this issue, we decompose the message passing process in GNNs into two operations: propagation and transformation. We then conduct a comprehensive analysis and quantification of the influence correlations between labels in each operation. Building on these insights, we propose a novel model, Label Influence Propagation (LIP). Specifically, we construct a label influence graph based on the integrated label correlations. Then, we propagate high-order influences through this graph, dynamically adjusting the learning process by amplifying labels with positive contributions and mitigating those with negative influence. Finally, our framework is evaluated on comprehensive benchmark datasets, consistently outperforming SOTA methods across various settings, demonstrating its effectiveness on MLNC tasks.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Multi-label node classification | DBLP (6:2:2) | Macro F1 Score86.5 | 12 | |
| Multi-label node classification | Delve-M (6:2:2) | Macro F158.87 | 12 | |
| Multi-label node classification | BlogCat (6:2:2) | Macro F159.02 | 12 | |
| Multi-label node classification | Abnomal (6:2:2) | Macro F153.98 | 12 | |
| Multi-label node classification | DBLP node split (6:2:2) | Macro AUC94.38 | 11 | |
| Multi-label node classification | BlogCat node split (6:2:2) | Macro AUC70.21 | 11 | |
| Multi-label node classification | OGB-p node split (6:2:2) | Macro AUC74.82 | 11 | |
| Multi-label node classification | PCG node split (6:2:2) | Macro AUC65.73 | 11 | |
| Multi-label node classification | HumLoc node split (6:2:2) | Macro AUC73.22 | 11 | |
| Multi-label node classification | EukLoc node split (6:2:2) | Macro AUC72.92 | 11 |