Design And Implementation Of Spatial Crowdsourcing System Supporting Task Allocation In Uncertain Environment

With the popularity of mobile devices,the technological progress of mobile phones and the improvement of wireless network bandwidth,the concept and mode of spatial crowdsourcing are widely used in the fields of intelligent transportation,environmental monitoring,social services and so on.Therefore,there are many spatial crowdsourcing systems serving various fields on the network.The spatial crowdsourcing system provides a platform for crowdsourcing businesses that publish tasks and crowdsourcing workers participating in tasks on the Internet,which greatly improves the transaction efficiency and income.Therefore,this paper designs and implements a spatial crowdsourcing system to support task allocation in an uncertain environment,which realizes five functions:personal management,task management,geographic information,result management and progress management,and meets the basic needs of spatial crowdsourcing activities.The personal management module includes user login and personal information,so that the system can effectively manage members participating in crowdsourcing activities;The task management module includes task release and task allocation,which enables the system to ensure the release and allocation of tasks;The geographic information module includes geographic information display and the shortest path,so that the system provides users with geographic data support;The progress management module includes progress query and update,progress control and message message,so that the system can comprehensively control the task progress;The result management module includes result evaluation and reward distribution,so that the system can better ensure the interests of crowdsourcing workers.Among them,the core problem of spatial crowdsourcing system is task allocation.Therefore,in the implementation of system task allocation module,this paper adopts the task allocation algorithm proposed in this paper to support uncertain environment.The task allocation algorithm in uncertain environment can realize task allocation in uncertain environment,that is,when the quality of workers is unknown.Firstly,this paper uses Markov decision process(MDP)in reinforcement learning to simulate spatial crowdsourcing problem in uncertain environment.Then,this paper uses sarsa algorithm to solve the spatial crowdsourcing task allocation problem in uncertain environment,and puts forward dynamic exploration method and comprehensive return function to improve sarsa algorithm.Finally,a large number of experiments show the effectiveness of the improved algorithm on real data sets.The spatial crowdsourcing system designed and implemented in this paper supports task allocation in uncertain environment,which meets the basic needs of spatial crowdsourcing activities.Among them,the task allocation algorithm in the uncertain environment used by the task allocation module ensures that when the quality of workers is unknown,it can effectively allocate tasks to appropriate crowdsourcing workers,so that crowdsourcing workers,crowdsourcing businesses and crowdsourcing platforms can improve revenue,which has high research value and application significance.