ValidatorRep: Blockchain-based Trust Management for Ensuring Accountability in Crowdsourcing

Crowdsourcing as a computing paradigm, has been widely used in industries and services. Accountability in crowdsourcing services enables participants to work honestly and improves the quality of services. The realization of accountability requires trusted evidence, multiparty verification, and fair reward or punishment. Blockchain technology, which is inherently tamper-resistant, traceable, and decentralized, puts forward a direction for realizing the requirements. However, in the process of data management and decentralized verification, it is hard to achieve a better trade-off between efficiency and security. This paper proposes a blockchain-based verification scheme integrated by trust management, 'validatorRep', that is suitable to enhance accountability in the crowdsourcing system. In detail, a decoupled blockchain model is proposed for the differentiated storage of business transactions and log transactions during data interaction. Additionally, a fine-grained trust model is proposed, including both the reputation of participants and the trust relationship between participants. Based on fine-grained trust, the decentralized verification scheme is designed to guarantee secure data access, trusted verification, and fair reward or punishment. Finally, the proposed framework is deployed on the Ethereum platform to observe its effectiveness and overall performance. Simulation results also reveal that the proposed fine-grained trust model can provide efficient accountability for crowdsourcing.