Research On Task Allocation Method Of Human-machine Cooperation In Group Computing

The rapid development of the Internet has promoted the emergence of big data,and people have paid more and more attention to the issue of the utilization of big data.However,the large scale,diversity,high speed,and low value density of big data have hindered the use of big data.In the face of these obstacles,the performance of a single computer can no longer meet the demand,and the concept of group computing for human-machine collaboration is correspondingly born.The group computing of humanmachine collaboration,which is a computing model in which people and cluster collaborate with each other in a big data environment,integrates Internet users' complex cognitive reasoning capabilities and computer clusters' high-performance computing capabilities to handle complex tasks that are difficult for computers to accomplish.This paper does some research on the two parts of human-machine collaboration,including the game method based on trusted users,and the matching method between the adaptively segmented task and the trusted user.1.In task assignment,there is a problem that tasks finished are uneven by Internet users,whose learning background and professional abilities are different.This paper proposes a game method based on trusted users.This method uses a dynamic Multi-domain cooperative allocation(MDCA),which mainly includes two steps: 1)First,the primary trusted users are screened through the evaluation of credibility;2)These trusted users play the game for specific task areas.Through experiments,it is shown that: 1)After the game,trusted users can better match tasks in a multi-domain range and have a high degree of relevance to the field in which the task is located;2)When there is a great degree of correlation between the user's professional field and the task,the quality and efficiency of the task completion can be improved,and the issue of alliance revenue and cost allocation can be better resolved through the game.2.In order to make the trusted Internet users after the game better complete the relevant tasks in a specific field and improve the quality of the specific subtasks,this paper proposes a method for bigraph matching of trusted users—specific task.This method uses the Hopcroft-Karp algorithm which is improved based on the Hungarian algorithm and matches the trusted crowd after the game with the specific subtasks after adaptive segmentation.The task random assignment algorithm and Hungarian algorithm are used as the experimental reference objects.This paper experimentally verifies that the advantages of Hopcroft-Karp algorithm compared with the other two algorithms are as follows: 1)Using specific data sets for testing,Hopcroft-Karp algorithm has better matching effect;2)Tests using data of different degree of difficulty and the same kind of degree of difficulty,users using the Hopcroft-Karp algorithm have higher accuracy in completing tasks.