Abstract: Abstract:The process of scientific research activities have a progressive, derived feature of complex structure.The scientific research innovation chain-data chain-publishing chain are closely connected and iterative.The evolution of data ecosystem and its model evolution is crucial to promoting scientific research innovation and expanding knowledge dissemination. On the basis of conceptual backtracking, this paper uses the theory of data ecosystem to elaborate the different data-driven logics of scientific research innovation and academic publishing. According to the different data thinking, data institution,data subject relationship, data management structure, data circulation environment and data management methods, the research data ecosystem can be divided into three gradual stages: closed, expanded and collaborative. Summarize the development trend of research data ecosystem from single to ecological, from fragmentation to system, from unilateral to collaborative, from point-to-point to integration, from manual to intelligent. A theoretical model for the evolution of scientific research data ecological model is proposed, and provide a new idea for promoting the overall quality improvement of each element and link of scientific research ecosystem with scientific data ecological governance.

摘要： 科学研究活动过程具有动态递进、扩张衍生的复杂结构性特征，科研创新链-数据链-出版链之间紧密衔接、循环迭代，数据生态及其模式演进对于促进科研创新、扩大知识传播至关重要。在概念回溯的基础上，运用数据生态理论深入阐述科研创新与学术出版的不同数据驱动逻辑。同时，从数据思维、数据制度、数据主体关系、数据管理结构、数据流通环境、数据管理方法的不同维度将科研数据生态划分为封闭型、扩张型、协同型三个演进阶段，总结科研数据生态从单一到融合、从碎片到系统、从单边到协同、从点对点到一体化、从手工到智能的发展趋势与表现形态，提出科学研究数据生态模式演进的理论模型，为促进科研生态体系各要素各环节的整体质量跃升提供了新的解决思路。