Crowdsourcee Evaluation Based on Persuasion Game

Kun Li,Shengling Wang,Xiuzhen Cheng

DOI: https://doi.org/10.1016/j.comnet.2019.04.019

IF: 5.493

2019-01-01

Computer Networks

Abstract:Cyber-physical systems (CPS) can integrate computational and physical capabilities so that people can interact with their environments in real time, forming the basis of smart technologies and hence improving our daily life. Due to the conflict between the high complexity that a CPS job often has and the weak capacity of a CPS node, it is wise to employ crowdsourcing mode making multiple CPS nodes can work collaboratively to complete difficult CPS job. Crowdsourcee evaluation is a critical technology in crowdsourcing to guarantee the quality of submitted tasks. The state-of-the-art crowdsourcee evaluation approaches have an underlying assumption: the evaluator is fair so that his evaluation is unbias. However, if an evaluator is a stakeholder, he or she may submit an unjust crowdsourcee evaluation report to the crowdsoucer to affect the final decision for maximizing his or her payoff. To solve this issue, we take advantage of the persuasion game to describe the relationship between the crowdsourcer and any evaluator, through which the optimal strategy for the crowdsourcer can be obtained to force the evaluator to reveal the actual ability of a crowdsoucee candidate. Such optimal strategies are deduced in both common settings as well as in more complicated scenarios. To the best of our knowledge, our work is the first one which solves the bias crowdsourcee evaluation issue resulting from the interest relevance of evaluators.