Secure and Reliable Decentralized Truth Discovery Using Blockchain

Yifan Tian, Jiawei Yuan, Houbing Song

Abstract: Truth discovery with crowdsourcing has become increasingly popular in recent years by leveraging the wisdom of crowd to solve complex tasks. So far, many existing crowd-sourcing applications utilize a central server for deployment, which collects and processes data from a crowd of workers. However, this kind of centralized model also introduces security concerns, including data privacy, integrity of processed data, and single point of failure. In this paper, we propose a secure decentralized framework for truth discovery with a privacy-preserving and reliable realization. Instead of relying on the central servers (e.g., public cloud servers)to behave correctly, our framework delegates the data aggregation and processing tasks to distributed entities, whose behaviors are enforced and verified by utilizing the emerging blockchain-based smart contract technology. Meanwhile, as the blockchain lacks support for on-chain data confidentiality, we utilize the privacy-preserving solution and integrate it with blockchain for privacy protection. Moreover, given the decentralized nature of our framework, it also overcomes the limitation of single point of failure. We implement a prototype of our framework using Ethereum blockchain and demonstrate its practical performance.