Modelling And Simulation For Detecting Vulnerabilities And Security Threats Of Smart Contracts Using Machine Learning

Abstract

Recently, the use and development of a blockchain systems such as Ethereum has increased rapidly, and many systems have relied on a third party as an intermediary between the sender and the receiver. Despite the attempts of developers to protect smart contracts, smart contracts contain many vulner-abilities that hackers resort to exploiting and using due to the attack that caused many financial and economic losses, and with the increase of errors in smart contracts, there are many tools and methods. For the analysis of smart contracts, machine learning models have appeared that facilitate their discovery instead of extracting them manually. In this paper, We have built a model that attempts to cancel the third party and we used machine learning to identify valid and invalid smart contracts. We have used several models and compared them with previous results of previous work in the same field. The result of this research was as expected of height accuracy achieved with approximately.99%.