基于三方演化博弈的众测激励机制研究

Research on the Crowd Testing Incentive Mechanism Based on Tripartite Evolutionary Games

摘要: 针对众测平台的管理问题，本文构建了平台方、发包方以及接包方的三方演化博弈模型及其复制动态方程，并对其均衡点的局部稳定性进行了分析，得到了三方在不同条件下的稳定性策略，进而研究声誉奖惩机制和补偿机制对完善众测平台管理以及参与各方策略选择的影响，并通过数值仿真进行验证。研究结果表明：单方的演化策略受到其他两方策略的影响；监管成本、声誉的正效应以及对接包方投机行为的惩罚力度直接影响着平台方的策略选择；预期收益以及平台方对接包方投机行为的惩罚力度直接影响着发包方的策略选择；声誉的正向激励以及惩罚力度直接影响着接包方的策略选择。本文为众测平台制定管理策略提供一定的理论指导和实践支持。

Abstract: Aiming at the management issues of crowdsourcing platforms, this paper constructs a tripartite evolutionary game model and its replicator dynamics equations for the platform operator, task requester, and task performer. We analyze the local stability of the equilibrium points and derive the stability strategies for the three parties under different conditions. Furthermore, we investigate the impact of reputation reward and punishment mechanisms as well as compensation mechanisms on improving the management of crowdsourcing platforms and the strategy choices of the participants, and verify these through numerical simulations. The research results indicate that the evolutionary strategy of one party is influenced by the strategies of the other two parties. The regulatory cost, the positive effect of reputation, and the punishment intensity for the task performer’s opportunistic behavior directly affect the strategy choice of the platform operator. The expected payoff and the punishment intensity for the task performer’s opportunistic behavior directly influence the strategy choice of the task requester. The positive incentive of reputation and the punishment intensity directly impact the strategy choice of the task performer. This paper provides theoretical guidance and practical support for the formulation of management strategies for crowdsourcing platforms.