Skip to content

Latest commit

 

History

History
44 lines (35 loc) · 1.74 KB

detecting-algthm.md

File metadata and controls

44 lines (35 loc) · 1.74 KB

de-faker-algthm

Purpose: Planning for the algorithm of detecting self-trading

Scenario #1

An abuser generates fake transactions through numerous wallets to avoid fake detection. Especially, creating new wallet does not take any cost in Ethereum, so cost-free abusing is available.

Detection 1. Number of interacting contracts from a wallet

For every wallet, count the number of contract which interact with the wallet. If the number approaches 1, acknowledge as a fake tx.

  • Requirement:
    • list of active wallets
    • list of interacted contract for active wallets
  • Output:
    • Average number of interacting contract
    • list of fake wallets

Scenario #2

An abuser owning a specific group of wallets utilizes those wallets to generate fake transactions. If the number of wallet is limited due to the wallet creating cost or technical reason, then this scenario might work.

Detection 1. Closed token flow

If we succeed in finding the token flow in a closed group of wallets, then it can be verified as a group fake transaction.

  • Requirement:
    • list of transactions of active wallets and interacting wallets
    • Closed group detecting algorithm
  • Output:
    • list of fake wallets in the closed group

Scenario #3

An abuser might use a same method repeatedly to generate fake transactions.

  • Temporal correlation
  • Transaction volume

Detection 1. Repeated transaction pattern

If we can find any correlations between transactions, then those transactions could be regarded a fake.

  • Requirement:
    • list of transactions from the contract
    • Pattern recognition algorithm (hopefully machine learning based?)
  • Output:
    • Repeated pattern and the fake tx generating scheme
    • list of fake transactions

Scenario #4

Detection 1. Liquid token cap