随着管理实践问题的日趋复杂，基于现代信息技术的大群体决策已成为近年来备受关注的重点研究课题。为了解决现有大群体决策方法中并未考虑专家知识结构影响的问题，首先从领域性、不完备性、可靠性三个视角对该类决策问题的性质和特征进行了描述和界定，并应用基本信任分配函数提出了不完备性推断信息的提取机理，然后以证据推理规则为技术基础，分别结合补偿性和非补偿性融合策略提出了对专家个体和专家群体的推断信息进行融合的方法和定理，在此基础上遵循先个体融合后群体融合的思路构建了考虑专家知识结构的大群体决策方法步骤，最后应用数值对比分析验证了提出方法的科学性和有效性。创新之处不仅体现在推断信息提取机理能够充分反映专家知识结构的影响，而且还体现在融合方法考虑了属性权重和专家可靠性的性质差异，可以反映属性信息之间的补偿性以及专家信息之间的非补偿性。这为提高和保证大群体决策的效率和效果将起到重要支撑作用。
With management practice problems becoming more and more complex, the large-group decision making based on modern information technology has been a noteworthy research topic recently. The large-group decision making may be influenced by expert's knowledge structure,which is not considered in existing methods. In order to solve above problem, the properties and characteristics are described and defined from three perspectives such as territoriality, incompleteness and reliability, and the mechanism for extracting incomplete inference information is presented by applying basic belief assignment function. Then the methods and theorems are constructed for fusing inference information of individual expert and expert's group respectively with compensatory strategy and non-compensatory strategy, both of which are technologically based on evidence reasoning rule. The procedure of multiple attribute large-group decision-making method with incomplete information by considering expert's knowledge structure is proposed by following the fusion thought that "from individual fusion to group fusion". The numerical comparison analysis shows the proposed method is scientific and effective finally. The innovation points of the proposed method are not only to reflect the influences by expert's knowledge structure within the mechanism for extracting incomplete inference information, but also to distinguish the differences between expert reliabilities and attribute weights and reflect the compensatory among attributes and the non-compensatory among experts. This will play an important supporting role in improving and ensuring the efficiency and effectiveness of large group decision making.
[1] Zahir S. Clusters in a group:Decision making in the vector space formulation of the analytic hierarchy process[J]. European Journal of Operational Research,1999, 112(3):620-634.
[2] 陈晓红,刘蓉. 改进的聚类算法及在复杂大群体决策中的应用[J]. 系统工程与电子技术,2006, 28(11):1695-1699.
[3] 徐选华,范永峰. 改进的蚁群聚类算法及在多属性大群体决策中的应用[J]. 系统工程与电子技术,2011, 33(2):346-349.
[4] Liu Bingsheng, Shen Yinghua, Chen Yuan, et al. A two-layer weight determination method for complex multi-attribute large-group decision-making experts in a linguistic environment[J]. Information Fusion,2015, 23:156-165.
[5] 田晓娟,王利东. 基于AFS理论的大群体决策中决策者权重确定方法[J]. 科学技术与工程,2015, 15(15):61-65.
[6] 张晓,樊治平. 一种基于证据推理的多指标多标度大群体决策方法[J]. 运筹与管理,2011, 20(2):15-20, 27.
[7] 张晓,樊治平. 基于前景随机占优的多属性多标度大群体决策方法[J]. 控制与决策,2014, 29(8):1429-1433.
[8] 张晓,樊治平. 一种基于随机占优准则的多指标多标度大群体决策方法[J]. 系统工程,2010, 28(2):24-29.
[9] Liu Yang, Fan Zhiping, Zhang Xiao. A method for large group decision-making based on evaluation information provided by participators from multiple groups[J]. Information Fusion,2016, 29:132-141.
[10] Xu Xuanhua, Zhong Xiangyu, Chen Xiaohong, et al. A dynamical consensus method based on exit-delegation mechanism for large group emergency decision making[J]. Knowledge-Based Systems, 2015, 86:237-249.
[11] Liu Bingsheng, Shen Yinghua, Chen Xiaohong, et al. A complex multi-attribute large-group PLS decision-making method in the interval-valued intuitionistic fuzzy environment[J]. Applied Mathematical Modelling, 2014, 38(17-18):4512-4527.
[12] 徐选华,万奇锋,陈晓红,等. 一种基于区间直觉梯形模糊数偏好的大群体决策冲突测度研究[J]. 中国管理科学, 2014, 22(8):115-122.
[13] 江文奇,丁健美. 基于参考点的大群体信息融合方法[J]. 系统工程, 2013, 31(11):8-12.
[14] Jiang Yuan, Xu Zeshui, Yu Xiaohan. Group decision making based on incomplete intuitionistic multiplicative preference relations[J]. Information Sciences, 2015, 295:33-52.
[15] 李德毅,杜鹢. 不确定性人工智能[M]. 北京:国防工业出版社, 2005.
[16] Yang Jianbo, Xu Dongling. Evidential reasoning rule for evidence combination[J]. Artificial Intelligence,2013, 205:1-29.
[17] Fu Chao, Chin K. Robust evidential reasoning approach with unknown attribute weights[J]. Knowledge-Based Systems, 2014, 59:9-20.
[18] Pipes R. Property and freedom[M]. New York:Knopf Doubleday, 1999.
[19] 杜元伟,杨娜. 大数据环境下双层分布式融合决策方法[J]. 中国管理科学, 2016, 24(5):144-154.
[20] 杜元伟,段万春,黄庆华,等. 基于头脑风暴原则的主观证据融合决策方法[J]. 中国管理科学, 2015, 23(3):130-140.
[21] 徐选华,王敏赛,陈晓红. 偏好冲突优化的多属性多阶段大群体决策方法[J]. 系统工程学报, 2014, 29(1):6.
[22] Chin K, Fu Chao, Wang Yingming. A method of determining attribute weights in evidential reasoning approach based on incompatibility among attributes[J]. Computers & Industrial Engineering,2015, 87:150-162.
[23] Kong Guilan, Xu Dongling, Yang Jianbo, et al. Combined medical quality assessment using the evidential reasoning approach[J]. Expert Systems with Applications, 2015, 42(13):5522-5530.
[24] Roselló L, Sánchez M, Agell N, et al. Using consensus and distances between generalized multi-attribute linguistic assessments for group decision-making[J]. Information Fusion, 2014, 17(0):83-92.
[25] Mkrtchyan L, Podofillini L, Dang V N. Bayesian belief networks for human reliability analysis:A review of applications and gaps[J]. Reliability Engineering & System Safety, 2015, 139:1-16.
[26] Smets P. Combination of evidence in the transferable belief model[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1990, 12(5):447-458.
