需求导向的综合评价集成框架及随机聚合求解

Although the research outputs in the field of comprehensive evaluation are very fruitful, some defects still exist, three of which are the focuses of this project including non-consistency of demand expression, conclusion non-consistency arising from multi-methods and absolutization of evaluation conclusions. Among the three defects, conclusion non-consistency arising from multi-methods is the biggest challenge, and its solution highly relies on the deep analysis of other two defects. Aiming at non-consistency of demand expression and absolutization of evaluation conclusions, the improving thoughts of development from method-oriented evaluation mode to demand-oriented evaluation mode, and further to demand-oriented integrated evaluation mode is addressed step by step in the project. In the face of complex issues, new mode addressed in the project may build an open, dynamic and constantly improving comprehensive evaluation integrated framework, and use comparatively perfect information and evaluation methods as well as stochastic clustered solution algorithms to get high quality conclusions following the rules of objective reality, so that the puzzle leading by conclusion non-consistency arising from multi-methods can be solved by the conclusions feedback, demand interaction and information update in the precession systematically. Finally, the new mode will be used in the measurement of Chinese province and city real-time performance in order to supply the foundation of concept of innovation, harmony, green, opening and sharing.

综合评价领域成果丰硕，但仍存在一些不足，项目关注其中三个重要方面，即“需求表达不一性、评价结论绝对化、多方法结论非一致性”三个现象，其中“多方法结论非一致性”是综合评价推进过程中最大的难题，而对于其它两个现象的深入分析是解决该难题的重要途径。本项目针对“需求表达不一性”及“评价结论绝对化”现象，递进地提出由“方法导向的评价模式”向“需求导向的评价模式”，再向“需求导向集成的评价模式”改进的思路。在面向复杂问题时，新模式可建立起开放、动态、不断旋进完善的综合评价集成框架，运用尽可能完整的信息、评价方法及随机聚合求解算法，形成遵循事物客观事实的高质量评价结论，从而使得评价需求者对于“多方法结论非一致性”的困惑在通过结论反馈、需求交互、信息更新的不断旋进过程中得到系统的解决。最后，将研究应用于“中国省域及城市实时绩效测度”问题中，为“创新、协调、绿色、开放、共享”这一新理念的持续推进提供支撑。

项目面向评价中“需求表达不一性、评价结论绝对化、多方法结论非一致性”等问题，提出了需求导向集成的评价模式，建立了开放、动态、不断旋进的综合评价集成框架，方便多来源、多结构、多关系、多偏好特征的评价信息的统一融合。随机聚合求解算法的研究为综合评价集成框架的整体求解提供了技术及方法支撑，形成优胜度矩阵统计被评价对象之间的优劣概率。基于此，构建富评价结论集，不仅包含体现被评价对象相对优劣的可能性排序结论，而且包含体现被评价对象发展动态的序列分值。依托项目理论研究成果，解决省域科技创新绩效、城市可持续发展绩效、及东北老工业基地振兴绩效评价等问题，助力省域、城市发展及东北振兴。依托项目研究内容，发表学术论文34篇，其中第一作者及通讯作者论文18篇、SCI、SSCI检索论文17篇。以第一作者在经济管理出版社出版学术专著《综合评价理论与方法（第二版）》1部。. 项目研究成果解决了“需求表达不一致性”及“评价结论绝对化”的传统评价现象，拓展了综合评价理论体系；多评价方案融合及随机聚合求解问题的研究提升了综合评价面向复杂信息及问题的能力，拓宽了综合评价解决实际问题的技术途径。应用研究成果可为助力东北老工业基地全面振兴贡献力量，并可为中国省域及城市的科技创新、可持续发展提供参考建议。