摘要
为提高移动众包系统的有效性和可靠性,设计了一套完整的在线激励机制优化算法,针对用户到达和参与任务的异步行为,提出一种改进的多阶段反向拍卖算法,通过在线学习自适应确定密度阈值,动态选择最优用户集,并在每次交易后对用户的信誉进行更新,以指导下次任务分配。仿真结果表明,该优化算法满足计算有效性、利益双方正收益性和真实性,能在一定预算和时间约束下获得更好的性能。
In order to improve the validity and reliability of mobile crowdsourcing system,a complete optimization algorithm of online incentive mechanism was designed. For the asynchronous behavior of users arriving and participating in tasks,an improved multi-stage reverse auction algorithm was proposed. Through online learning adaptively determine the density threshold,dynamically selected the optimal user set,and updated the user’s reputation after each transaction to guide the next task assignment. The simulation results show that the optimization algorithm can meet the computational efficiency and the profitability and authenticity of both parties,and can achieve better performance under certain budget and time constraints.
作者
张永棠
Zhang Yongtang(Dept. of Computer Science & Technology,Guangdong Neusoft Institute,FoshanGuangdong 528225,China;Guangdong Key Laboratory of Big Data Analysis &Processing,Guangzhou 510006,China)
出处
《计算机应用研究》
CSCD
北大核心
2019年第9期2588-2589,2595,共3页
Application Research of Computers
基金
国家自然科学基金资助项目(61363047)
广东省大数据分析与处理重点实验室开放基金资助项目(2017007)
佛山市科技创新项目(2016AG100792)
关键词
移动众包
数据感知
优化算法
智能优化
mobile crowdsourcing
data awareness
optimization algorithm
intelligent optimization
