群体计算中基于博弈论的任务分配策略

Task Allocation Strategy Based on Game Theoryin Group Computing

大数据的快速发展,推动了社会经济和科技的发展,但大数据的价值密度低等特点为其发展带来了挑战。大数据的这些特点使得大数据迫切需要复杂认知的推理技术,而人机协作的群体计算成为了复杂认知推理技术的有效途径,但其任务分配策略还尚未完善。尽管已经有学者提出了基于用户主题感知的任务分配策略,解决了涉及不同专业背景及不同知识水平的任务分配,但并未解决处于同层次知识水平和专业背景的用户如何分配任务,使得计算效率更高。针对此问题,提出了基于博弈论的任务分配算法,检测相同专业背景和知识水平的人群完成任务的准确率,与任务随机分配相比较,突出博弈论算法的准确性。

The rapid development of big data promotes the progress of economy and technology,but the features,low value density,bring challenges to the big data.These features make that it need urgently complex cognitive reasoning,which is effectively solved by human-machine collaboration based crowd computing.However,crowd computing＇s task allocation strategy is not maturity completely.some scholars have come up with a theme-aware task assignment framework,which solves task allocation to different professional background and knowledge level,but it does not deal with task allocation involving same knowledge level and professional background,which makes higher computational efficiency.To deal with this problem,it propose a task allocation algorithm based on game theory,which detects the accuracy with same professional and knowledge level.The task allocation algorithm based on game theory,compared with randomly task allocation,shows the accuracy of game theory algorithm.