Research On Task Assignment Of Mobile Crowdsourcing In Crowd Intelligence Network

In the era of the Internet of Everything,with the continuous advancement of intelligence.The number of crowd users is increasing exponentially.The data throughput of crowdsourcing also shows a trend of hundreds of millions of growths.While the application and research of spatio-temporal crowdsourcing brings opportunities,the application and research of spatio-temporal crowdsourcing also poses huge challenges.Spatio-temporal crowdsourcing also contains a lot of different research content.One of the core issues is task assignment.Existing task assignment algorithms can be mainly divided into offline and online.Most of the existing task assignment research focuses on offline task assignment.It is assumed that the platform has already learned all the information about crowd workers and crowd tasks before the assignment.However,due to the randomness of crowd workers and crowd tasks,it is difficult for these studies to get good results in real scenes.Although the online algorithm solves the problem of the offline algorithm’s poor performance in the online model,its optimization content is also more difficult and complicated.To balance the assignment amount and utility of online task assignment under spatio-temporal crowdsourcing,the research content of this paper mainly includes:(1)An online bilateral task assignment(OBA)problem is proposed for the online model.The competition ratio of the greedy algorithm under the OBA problem model is analyzed.Based on the greedy algorithm,another solution to the OBA problem is proposed and named as the improved baseline algorithm.Aiming at the success rate of workers in the OBA problem model,Two-stages Bilateral Online Priority Reassignment Algorithm(BOPR)is proposed.In order to ensure the number of task matching,a priority queue is designed in the BOPR algorithm,the priority is sorted by considering the time attributes of tasks and workers.Design a two-stages assignment strategy to improve the success rate of task assignment.(2)A cross-regional online task assignment model(CROT)is proposed for the online model and the cross-regional movement of workers.Based on the CROT model,an offlineguided cross-regional task assignment algorithm(OTARP)is proposed.By the idea of edge cloud to complete the assignment in two ways: intra-regional and inter-regional.The first stage uses historical data for offline prediction.The second stage uses the offline data of the first stage to guide the online task assignment process,speeds up the task assignment process through multiple rounds of assignment,optimizes the assignment process by combining offline guidance and online assignment.In order to encourage workers to complete tasks across regions,a reasonable incentive strategy is designed to motivate workers to accept tasks across regions.To help workers match more tasks,a Drop-By Rider Model(DBR)is designed to optimize the number of assignments and improve utility.