CreditXAI: A Multi-Agent System for Explainable Corporate Credit Rating

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Shi, Yumeng, Yang, Zhongliang, Wang, Yisi, Zhou, Linna
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908611890905088
author Shi, Yumeng
Yang, Zhongliang
Wang, Yisi
Zhou, Linna
author_facet Shi, Yumeng
Yang, Zhongliang
Wang, Yisi
Zhou, Linna
contents In the domain of corporate credit rating, traditional deep learning methods have improved predictive accuracy but still suffer from the inherent 'black-box' problem and limited interpretability. While incorporating non-financial information enriches the data and provides partial interpretability, the models still lack hierarchical reasoning mechanisms, limiting their comprehensive analytical capabilities. To address these challenges, we propose CreditXAI, a Multi-Agent System (MAS) framework that simulates the collaborative decision-making process of professional credit analysts. The framework focuses on business, financial, and governance risk dimensions to generate consistent and interpretable credit assessments. Experimental results demonstrate that multi-agent collaboration improves predictive accuracy by more than 7% over the best single-agent baseline, confirming its significant synergistic advantage in corporate credit risk evaluation. This study provides a new technical pathway to build intelligent and interpretable credit rating models.
format Preprint
id arxiv_https___arxiv_org_abs_2510_22222
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CreditXAI: A Multi-Agent System for Explainable Corporate Credit Rating
Shi, Yumeng
Yang, Zhongliang
Wang, Yisi
Zhou, Linna
Multiagent Systems
Computational Engineering, Finance, and Science
In the domain of corporate credit rating, traditional deep learning methods have improved predictive accuracy but still suffer from the inherent 'black-box' problem and limited interpretability. While incorporating non-financial information enriches the data and provides partial interpretability, the models still lack hierarchical reasoning mechanisms, limiting their comprehensive analytical capabilities. To address these challenges, we propose CreditXAI, a Multi-Agent System (MAS) framework that simulates the collaborative decision-making process of professional credit analysts. The framework focuses on business, financial, and governance risk dimensions to generate consistent and interpretable credit assessments. Experimental results demonstrate that multi-agent collaboration improves predictive accuracy by more than 7% over the best single-agent baseline, confirming its significant synergistic advantage in corporate credit risk evaluation. This study provides a new technical pathway to build intelligent and interpretable credit rating models.
title CreditXAI: A Multi-Agent System for Explainable Corporate Credit Rating
topic Multiagent Systems
Computational Engineering, Finance, and Science
url https://arxiv.org/abs/2510.22222