Towards fair and efficient task allocation in blockchain-based crowdsourcing
- PDF / 1,533,088 Bytes
- 12 Pages / 595.276 x 790.866 pts Page_size
- 66 Downloads / 210 Views
REGULAR PAPER
Towards fair and efficient task allocation in blockchain‑based crowdsourcing Xiaoyi Pang1 · Dengfeng Guo1 · Zhibo Wang1 · Peng Sun1 · Liqiang Zhang1 Received: 1 June 2020 / Accepted: 23 October 2020 © China Computer Federation (CCF) 2020
Abstract Crowdsourcing has been a popular paradigm leveraging the power of the crowd to accomplish a common goal. Traditional crowdsourcing systems rely on a centralized platform to allocate tasks and rewards to users, facing severe problems of single node failure and malicious behaviors of the platform. Recently, some works take advantage of blockchain to solve the drawback of centralization. However, the key issue, fair and efficient task allocation, has not been explored in existing blockchain-based crowdsourcing systems. In this paper, we design a novel blockchain-based framework for crowdsourcing, in which a distributed reverse and blind auction-based task allocation mechanism (RbatAlloc) is proposed utilizing user profile and bidding price to realize fair and efficient task allocation in transparent blockchain environment. Finally, we implement a prototype of our system and deploy it to a locally developed network. The experimental results demonstrate the effectiveness of the proposed framework and mechanism. Keywords Crowdsourcing · Decentralization · Blockchain · Task allocation
1 Introduction Crowdsourcing, leveraging the power of the crowd to complete tasks, has received great attentions in recent years. A typical crowdsourcing system consists of requesters, workers and a hosting platform. The requesters can post tasks on the platform with the task descriptions, requirements and rewards. Workers can apply for tasks they are interested in and get rewards when they complete tasks. In traditional crowdsourcing systems, the hosting platform is controlled by a centralized third party that serves as the brain keeps
* Zhibo Wang [email protected] Xiaoyi Pang [email protected] Dengfeng Guo [email protected] Peng Sun [email protected] Liqiang Zhang [email protected] 1
School of Cyber Science and Engineering, Wuhan University, Wuhan, People’s Republic of China
the private data of users and tasks and is responsible for the common operations of crowdsoucring (e.g. task allocation). However, the centralized platform is prone to be attacked or even seized, which is the so called “single node failure”. Once the platform is lost control, users will face problems of privacy leakage, information abuse, payment cheating, deposit loss and so on. For example, some personal sensitive information of users is usually required for the platform to assign tasks to a group of reliable and professional users. The required information, such as location, identity, job, hobbies, poses a threat to user life and property once the platform is under control and exposes user information to attackers. This will in turn affect users’ willingness to participate in crowdsourcing. Recently, the necessity of decentralization (Ren et al. 2017, 2019; Li et al. 2017b; Feng and Yan 2019; M
Data Loading...
