空间众包环境下的任务定价模型研究
2020-07-09郑军林蔓佳胡蓉
郑军 林蔓佳 胡蓉



摘 要 在综合考虑平台、商家和会员三方相互作用的基础上,通过建立包含任务动态分配机制的动态规划模型,结合金融定价思想刻画任务定价问题,并通过空间可视化对珠三角地区劳务众包平台数据进行实证研究.为提高模型的实用性,利用K-means聚类分析对任务打包并引入激励规则对动态定价模型进行了改进.最后,通过模拟仿真得出改进后模型的任务完成率为88.10%,相比平台现有定价模型(62.50%)和改进前的动态定价模型(85.20%)任务完成情况有较大幅度的提升.为基于地理位置的服务平台的商品定价、以及地理位置信息与平台会员的关系等实证和应用研究提供了理论与实践参考.
关键词 运筹学与控制论;动态任务定价;激励规则;动态规划
中图分类号 F224.3 文献标识码 A
Abstract Through conducting spatial visualization on the data from Crowdsourcing platform of the PRD region and applying financial pricing philosophy into Crowdsourcing mission scenarios, this paper established a dynamic programming pricing model with task automatic allocation scheme after comprehensive consideration among the task releasing entrepreneur, the receiver and the Crowdsourcing platform. In order to hone the model for actual application, the mission packing by K-means clustering and motivation rules improved on the dynamic programming pricing model. After solving the model and analog simulation, the latest dynamic model proved its effectiveness at the mission completion rate of 88.10%, while the rate of the present model at the platform was 62.50% and the pricing model before improvement was 85.20%. In conclusion, it provides much mentality for the location based service (LBS) platforms like express errands service, DiDi and Meituan takeout on quantifying the relation among service pricing, space location and members for both empirical study and application in reality.
Key words operational research and cybernetics; dynamic task pricing; incentive rules; dynamic programming
1 引 言
眾包(Crowdsourcing)是一种基于位置的服务,它利用手机中的GPS模块捕捉用户所处的位置并将位置信号传送到定位后台来实现空间定位.在双重定位核对系统下,其定位信息具有较高的精确度.它可以满足消费端的各种需求,也可以为厂商提供各种宣传、营销服务.现阶段国内外空间众包平台已经日益成熟,如美团外卖、跑腿服务和滴滴出行等.
目前,国内外对众包定价方法进行了研究,并取得了一些研究成果.张月蕾等(2018)[1]基于K-means聚类分析了影响任务定价的主要因素(任务经度、纬度、会员密度和会员信誉度),并以这些因素为自变量建立了线性定价模型.但该……
