DGP: A Dual-Granularity Prompting Framework for Fraud Detection with Graph-Enhanced LLMs (AAAI 2026)
Pending approval from our industry partner before public release.
If you use DGP in a scientific publication, we would appreciate citations to the following paper:
@article{li2025dgp,
title={DGP: A Dual-Granularity Prompting Framework for Fraud Detection with Graph-Enhanced LLMs},
author={Li, Yuan and Hu, Jun and Hooi, Bryan and He, Bingsheng and Chen, Cheng},
journal={arXiv preprint arXiv:2507.21653},
year={2025}
}