JS version of https://github.com/Bill0412/nft-wash-trading-detection

Wash Trading Cycle Detection

This project applies DFS to ERC721 NFT trading time-series transactions to find out the probability of a trader performing wash trading.

The time complexity for this project is O(E), while in reference 1, the algorithm to detect cycles is claimed to be O((N+E)(C+1)).

E: the number of edges in the graph, i.e., the number of transfer and trade.
N: the number of nodes in the graph, i.e., the number of users that bought the NFT.
C: the number of cycles in the graph.

Note: the sample data does not include the transfer data. So it only counts counts cycles before first transfer of an NFT.

Thie project is based on the following paper and improves it in terms of algorithmic time complexity.

NFT Wash Trading - Quantifying suspicious behaviour in NFT markets: https://arxiv.org/pdf/2202.03866.pdf