Data Description

This data is used in “Enhancing Ethereum Fraud Detection via Generative and Contrastive Self-supervision” for Ethereum fraud detection research. It is extracted and constructed from the Ethereum transaction records.

The data altogether extracts the two-hop neighbors of the target node and distinguishes between accounts and interaction types to form a heterogeneous Ethereum interaction graph (HEIG).

Download

Web

Reference

If you find this work useful, please cite the following:

@article{jin2024enhancing,
  title={Enhancing Ethereum Fraud Detection via Generative and Contrastive Self-supervision},
  author={Jin, Chenxiang and Zhou, Jiajun and Xie, Chenxuan and Yu, Shanqing and Xuan, Qi and Yang, Xiaoniu},
  journal={arXiv preprint arXiv:2408.00641},
  year={2024}
}