Stackelberg Game Based Tasks Assignment Mechanism Using Reputation in Crowdsourcing

Crowdsourcing is a new paradigm emerged in recentyears. It can deal with the tasks posted by the requestor, who wants the worker to accept the task and finish it. Incrowdsourcing, it is a common case that using the reputationmechanism estimates worker's ability to avoid that the workercontributes low-quality work. Therefore, the reputation mechanismis integrated into the crowdsourcing for the tasks assignmentand the biding price in this paper. However, it is undesirable forsome requestors that the evaluation for workers, as requestors'private information, is exposed. The challenge is to finish the tasksmentioned above with keeping the requestors' private informationfrom exposing. Another important challenge with insufficientattention resides in finding the communication channels and getthe necessary information, which can obtain optimal benefit inincentive mechanism, as most of researchers focus on competitiverelationship between the worker and requestor. In this paper, wepropose the novel framework using the reputation mechanismbased on the Stackelberg game model to focus on the cooperationbetween workers and requestors. There are two stages whereworkers and requestors observe each others' strategies or sharetheir information to each other to maximize their own benefit. Firstly, we formulate the framework based on the Stackelberggame model and discuss its advantage. Subsequently, we study theoptimal strategies of each, give the process that how to calculateit, and analyse the unique Stackelberg Equilibrium. Finally, wesimulate our the framework and use different numerical valueparameters to test the effect on the performance of the games.