概率语言术语集理论及其在智慧医疗健康服务中应用

Probabilistic linguistic term set is a brand-new form to express linguistic information in qualitative decision making, whose basic units are composed of a collection of linguistic terms and their corresponding probabilities. It can portray and handle linguistic preference information comprehensively and detailedly in the process of decision making, and thus, the probabilistic linguistic term sets have wide application prospects in various fields, such as medical management, personnel appraisal, investment decision making, service performance evaluation, etc. We will systematically investigate the probabilistic linguistic term set theory, and apply it into the field of wisdom medical treatment and health services, which include the following contents: (1) The project studies the basic operational laws of probabilistic linguistic term sets, and their aggregation operators and clustering methods; (2) The project studies the distance measures, similarity measures, correlation measures and entropy measures of probabilistic linguistic term sets; (3) The project studies the priority methods of probabilistic linguistic preference relations, the consistency checking criteria and the inconsistency improving algorithms of probabilistic linguistic preference relations, and then constructs the analytic hierarchy process and the analytic network process with feedback system based on probabilistic linguistic preference relations respectively; (4) The project studies the approaches for multiple attribute decision making and the theoretical mechanism for dynamic decision making based on probabilistic linguistic preference relations respectively; (5) The project establishes the theoretical framework of probabilistic linguistic decision making and the corresponding decision support system; (6) Based on the research results above, the project develops the corresponding application software packages to solve the important practical problems of medical diagnosis, medical resources allocation, and the construction of medical recommendation system in wisdom medical treatment and health services.

概率语言术语集是定性决策中一种崭新的语言信息表达形式，其基本单元由语言术语和相应的概率构成，在决策过程中可以全面细致地描述和处理决策者的语言偏好信息。它在医疗管理、人事考核、投资决策、服务绩效评估等诸多领域有着广阔的实际应用前景。本项目将系统地研究概率语言术语集理论及其在智慧医疗健康服务中应用。具体内容包括：研究概率语言术语集的基本运算法则、融合方式与聚类方法；研究概率语言术语集的距离测度、相似性测度、关联测度和熵测度等测度理论；研究概率语言偏好关系的排序方法、一致性检验准则和非一致性修正算法，分别建立基于概率语言偏好关系的层次分析法和具有反馈系统的网络分析法；研究基于概率语言术语集的多属性决策途径和动态决策机制；系统地构建基于概率语言信息的决策理论框架和决策支持系统；将上述研究成果研制成软件应用于智慧医疗健康服务中的医疗诊断、医疗资源合理配置、医疗推荐系统构建等重要实际问题。

概率语言术语集是一种新型的语言信息表达方式，它既能刻画决策者对客观事物不确定性的认知，又可以描述不同评价意见的重要性。在决策过程中，概率语言术语集作为一种切实可行且有效的工具可以全面细致地表达和处理决策者的语言偏好信息，从而很好地提升决策结果的合理性和可信度。目前，有关概率语言术语集理论和应用的研究受到了国内外相关领域学者的极大关注。项目组对概率语言术语集理论及其在多属性决策中的应用进行了深入而系统的研究，取得了一系列原创性研究成果，如：（1）丰富了概率语言术语集的概念，并对其运算规则、决策方法以及偏好关系的一致性进行了深入研究，同时融合了概率语言术语集的应用场景；（2）分析了概率语言信息环境下的信息集成、测度及聚类方法，进一步完善了概率语言信息融合理论，拓展了概率语言信息融合的新方法；（3）提出了概率语言判断矩阵的一系列排序方法，详细地研究了概率语言判断矩阵个体一致性和群体共识性，创建了概率语言层次分析法等一系列拓展方法；（4）建立了基于概率语言信息的群决策和多属性决策理论和方法，构建了基于概率语言信息的决策理论框架和决策支持系统；（5）在概率语言术语集的基础上进一步拓展，开创性地提出了概率犹豫模糊集，不确定概率语言术语集，双层犹豫模糊语言术语集，概率语言向量集与嵌套概率语言术语集；（6）对双层语言术语集和嵌套概率语言术语集的理论进行系统性地梳理和研究，同时丰富了双层语言术语集和嵌套概率语言术语集的应用研究。此外，对基于其它多种模糊信息（如：区间直觉模糊集、犹豫模糊语言术语集、双层犹豫模糊语言术语集、嵌套概率数值语言信息、概率偏好信息和语言修饰集等）的相关理论和决策方法也进行了深入地研究；把上述成果应用于智慧医疗服务、医疗分级诊断、公共卫生应急决策、海洋环境风险评估、地震紧急救援方案评估、以及水资源保护投资项目评估等诸多实际问题中。