WIRE: Web3 Integrated Reputation Engine

Abstract

Distributed Applications (DApps), powered by smart contracts, have sparked a significant transformation in the Web3 ecosystem by enabling the execution of real-world contracts on decentralized networks. However, the growing popularity of DApps has also led to an increase in malicious activities exploiting smart contracts, thereby exposing users to greater financial risks. Inspired by the FICO score system in traditional finance, we introduce WIRE, a reputation engine designed to evaluate the trustworthiness of deployed DApps. WIRE first derives diverse properties from contract activities, rather than relying solely on potentially irrelevant or unavailable source code. Based on selected properties, WIRE trains a machine learning model for assessing the trustworthiness of individual contracts. Further-more, WIRE utilizes a bytecode disassembler to identify related contracts of a DApp, thus determining its overall trustworthiness score. Moreover, WIRE's dashboard offers explainable and detailed reports that are accessible to users without professional knowledge. The evaluation results show that WIRE can provide a reliable and explainable reputation score for DApps. As a result, WIRE's users can distinguish between benign and malicious DApps or contracts with a high confidence.