科研大数据质量控制研究:角色定位、演化博弈与实施策略

An Analysis of Scientific Research Big Data Quality Control:Role Positioning,Evolution Game and Implementation Strategy

[目的/意义]数据污染、数据风险、数据壁垒等数据质量隐忧,影响着科研工作者的科技创新步伐,科研大数据质量控制等问题的研究迫在眉睫。[方法/过程]文章在对科研大数据质量控制困境进行分析的基础上,对科研大数据质量控制主体进行了角色定位,进而从收益和成本角度出发,构建了科研大数据质量控制的“三方”演化博弈模型,对参与博弈的科研大数据提供者、科研大数据消费者、科研大数据监管者的行为策略进行了博弈分析,并利用Matlab对三方的策略选择进行仿真,最后提出了科研大数据质量控制策略。[结果/结论]科研大数据质量控制演化博弈模型由科研大数据提供者、科研大数据消费者、科研大数据监管者三方主体构成,通过模型分析与仿真得出了博弈三方的演化稳定策略,提出了“基-翼”相济式优化策略“、响应式-联盟化”管控策略、全过程技术融合策略、人才-平台双驱联动策略四位一体的科研大数据质量控制实施策略体系。

[Purpose/significance]Data pollution,data risks,data barriers and other data quality concerns affect the pace of scientific innovation of researchers,research on issues such as quality control of scientific research big data is imminent,the state,the community are actively exploring research on the quality control of scientific research big data.[Method/process]The study shows that the evolutionary game model of research big data quality control is composed of three parties:research big data providers,research big data consumers and research big data regulators,and the evo⁃lutionary and stable strategies of the three parties of the game are derived through model analysis and simulation,and then the roles of research big data quality control are analyzed,and the principles of quality control are given,and final⁃ly the four-in-one research big data quality control implementation strategy system of"base-wing"phase optimization strategy,"responsive-alliance"control strategy,whole process technology integration strategy,and talent-platform du⁃al-drive linkage strategy is proposed.[Result/conclusion]The study shows that the evolutionary game model of scien⁃tific research data quality control is composed of three parties:scientific research data providers,scientific research da⁃ta consumers and scientific research data regulators,and the evolutionary stabilization strategies of the three parties are derived through model analysis and simulation.Furthermore,the role positioning of quality control subject of scientific research big data was analyzed.Finally,we propose a four-in-one research big data quality control strategy system,in⁃cluding"base-wing"optimization strategy,"responsive-alliance"control strategy,whole process technology integra⁃tion strategy,and talent-platform dual drive strategy.