Graph Dimension Attention Networks for Enterprise Credit Assessment

TL;DR

This paper introduces GDAN, a novel graph neural network with dimension-level attention for more accurate enterprise credit risk assessment, along with a new interpretability method and a real-world dataset.

Contribution

We propose GDAN, a new GNN architecture with dimension-level attention, and introduce GDAN-DistShift for interpretability, plus provide a real-world enterprise credit dataset.

Findings

GDAN outperforms existing methods on ECAD, SMEsD, and DBLP datasets.

GDAN-DistShift effectively explains message-passing distribution shifts.

The new dataset ECAD supports future research in enterprise credit assessment.

Abstract

Enterprise credit assessment is critical for evaluating financial risk, and Graph Neural Networks (GNNs), with their advanced capability to model inter-entity relationships, are a natural tool to get a deeper understanding of these financial networks. However, existing GNN-based methodologies predominantly emphasize entity-level attention mechanisms for contagion risk aggregation, often overlooking the heterogeneous importance of different feature dimensions, thus falling short in adequately modeling credit risk levels. To address this issue, we propose a novel architecture named Graph Dimension Attention Network (GDAN), which incorporates a dimension-level attention mechanism to capture fine-grained risk-related characteristics. Furthermore, we explore the interpretability of the GNN-based method in financial scenarios and propose a simple but effective data-centric explainer for GDAN, called GDAN-DistShift. DistShift provides edge-level interpretability by quantifying distribution shifts during the message-passing process. Moreover, we collected a real-world, multi-source Enterprise Credit Assessment Dataset (ECAD) and have made it accessible to the research community since high-quality datasets are lacking in this field. Extensive experiments conducted on ECAD demonstrate the effectiveness of our methods. In addition, we ran GDAN on the well-known datasets SMEsD and DBLP, also with excellent results.