IMPLEMENTASI DEEP LEARNING SEBAGAI UPAYA MENDETEKSI TINDAK PENIPUAN TOKEN KRIPTO PADA JARINGAN ETHEREUM

With the increasing popularity of blockchain technology and the use of crypto tokens, the problem of financial security in blockchain transaction networks have become more serious. There have been many cases of fraud in blockchain technology, especially on the Ethereum network. To overcome this problem, this research designs and builds a system that can detect fraudulent acts on ERC-20 tokens on the Ethereum network. The method used in this study is the Deep Neural Networks (DNN) algorithm to detect objects from scam tokens. The results of the performance evaluation show that the model created is capable of detecting fraud very well with a very high level of accuracy, which is above 99%. In testing the app, 100 new tokens were randomly purchased and resold without issue, indicating that the app can distinguish between scam tokens and non-scam tokens.

Keywords : Blockchain, Scam, Ethereum, Crypto Token, Deep Neural Networks