基于综合集成的中医药专家共识方法学研究

本项目属于系统科学、思维科学、信息科学等交叉学科在中医知识领域应用的热点研究，在前期工作的基础上，以自主研发的基于综合集成的中医信息评价系统为依托，针对中医药领域个体专家经验显性知识表达和隐性知识外化、群体专家意见收敛和共识中存在的多人多属性决策、多维信息不确定等问题，在综合集成法的指导下，运用“加权平均+动态调整权重+极化”等递进算法模型，设计构建基于贝叶斯推理判定的中医专家共识网络模型算法体系，完成由专家体系、知识体系、机器体系组成的基于综合集成的中医专家实时交互网络共识系统，在中医诊疗实践中探索验证，反复测试，使三个体系之间不断交互、反馈；完成人机结合、定性认识与定量分析相结合、专家挖掘与数据挖掘相结合，进而在互联网络平台上完成解决中医个体专家经验表达、群体专家意见聚合等信息处理问题，完成用概率推算、数值量化等表达专家共识并做科学决策的过程。

This project belongs to the hot spot research on the application of interdisciplinary subjects such as System Science, Thinking Science and Information Science in the field of TCM knowledge. On the issues of multi-attribute decision making problem and multi-dimensional information uncertainty problem in TCM consensus research，such as explicit knowledge expression and implicit knowledge externalization of individual expert experience, convergence of group expert opinion and consensus，We will use the progressive algorithm model of "weighted average + dynamically adjusted weight + polarization", such as building design based on Bayesian inference network model of traditional Chinese medicine experts consensus algorithm of decision system to design and construct a TCM expert consensus network model algorithm system based bayesian inference, and Complete the real-time interactive network consensus system of TCM experts based on Meta-synthetic composed of expert system, knowledge system and machine system, which took the Metasynthesis as the guidance, the TCM diagnosis and treatment support platform developed by ourselves as support. And then the system will be explored, verified and tested repeatedly in TCM diagnosis and treatment practice, which makes the three systems interact and feedback constantly. Man-machine combination, combining the qualitative knowledge and quantitative analysis, expert mining and the realization of the combination of data mining, make the system implementation of traditional Chinese medicine (TCM) on the Internet platform to solve individual expression of expert experience, group expert opinion aggregation information processing problems, such as, complete with probability prediction, numerical quantitative expression of expert group consensus and make scientific decision process.