Two-sided preferences task matching mechanisms for blockchain-based crowdsourcing. (1st October 2021)

• Record Type: Journal Article
• Authors: Kadadha, Maha; Otrok, Hadi; Singh, Shakti; Mizouni, Rabeb; Ouali, Anis
• Is Part Of: Journal of network and computer applications. Volume 191(2021)
• Publication Date: 2021-10-01
• Subjects: Two-sided preference -- Matching -- Blockchain -- Crowdsourcing -- Smart contract -- Gale–Shapley; Microcomputers -- Periodicals; Computer networks -- Periodicals; Application software -- Periodicals; Micro-ordinateurs -- Périodiques; Réseaux d'ordinateurs -- Périodiques; Logiciels d'application -- Périodiques; Application software; Computer networks; Microcomputers; Periodicals; 004.05; 004
• DOI: 10.1016/j.jnca.2021.103155
• Languages: English
• ISSNs: 1084-8045

Abstract

Abstract: In this paper, novel task matching mechanisms with two-sided preferences of workers and requesters are proposed for blockchain-based crowdsourcing. Existing blockchain-based crowdsourcing frameworks match workers to tasks using allocation mechanisms considering metrics such as cost, location, and workers' reputation to answer requesters' requirements. However, they still match workers to tasks through mechanisms that are requester-biased with no consideration for workers' preferences. This may lead to workers rejecting or neglecting their allocated tasks. As a solution, we propose two-sided preferences task matching mechanisms for blockchain-based crowdsourcing, namely SenseChain+, and Gale–Shapley Matching (GSM). In order to do the matching, the proposed mechanisms utilize the preferences of workers based on a proposed Quality-of-Task (QoT) metric, and the preferences of tasks based on a Quality-of-Information (QoI) metric. To ensure their autonomous, reliable, and transparent execution, these matching mechanisms are integrated in an existing blockchain-based crowdsourcing framework, implemented using smart contracts. The mechanisms and framework are implemented using Solidity on a private blockchain and evaluated using a real dataset. They are benchmarked to the Nearest Neighbor Matching (NNM) mechanism. The proposed mechanisms demonstrate higher performance compared to the benchmark in terms of workers' QoI, payment, satisfaction, and confidence. To demonstrate the need for each mechanism, the performance under different demand to supply contexts is measured in terms of the workers' QoI, confidence, and the minimum payment. Each proposed matching mechanism was found to outperform the others in a range of demand to supply ratios. Finally, the proposed matching mechanisms are stable and feasible on-chain with reasonable execution cost. Graphical abstract: