Dreamerryao / nft-wash-trading

An unofficial code implementation of the paper "NFT Wash Trading - Quantifying suspicious behaviour in NFT markets"

NFT Wash Trading

An unofficial code implementation of the paper "NFT Wash Trading Quantifying suspicious behaviour in NFT markets".

Each NFT is represented as a graph, with nodes representing addresses and edges representing transactions.

Records of transfers or sales form a directed graph.

The complexity of detecting and extracting cycles is O((N+E)(c+1)), where (N) represents the number of nodes, (E) represents the number of edges, and (c) represents the number of cycles in the graph.

The example data is powered by NFTGo. If you need more data, you can obtain an API-KEY by visiting this page, import it into the .env file, and run python crawl_nft_data.py -tx.

Usage

1. unzip nft_data.zip
2. run node detect.js

References

1. NFT Wash Trading - Quantifying suspicious behaviour in NFT markets: https://arxiv.org/pdf/2202.03866.pdf
2. Detect Cycle in Directed Graph Algorithm: https://www.youtube.com/watch?v=rKQaZuoUR4M