考虑专家信任风险行为挖掘与动态管理的社会网络群体共识决策研究

doi:10.16381/j.cnki.issn1003-207x.2022.2549

摘要/Abstract

摘要：

随着社会网络技术的发展，决策专家间的信任关系为群体共识决策研究带来了新的机遇和挑战。由于信任固有的风险性，专家在决策过程中可能存在被盲目信任或缺失应有信任的情况，进而产生影响决策的风险问题，导致达成科学可靠的共识更加困难。为此，本文依据专家行为挖掘，将社会网络群体共识决策中的信任风险与决策结果的科学可靠性相联系，提出考虑专家信任风险行为挖掘与动态管理的社会网络群体决策共识模型，以降低决策风险，提升决策结果的可信度与共识效率。首先，综合考虑社会网络与专家可信度，设计基于t-范数的专家风险行为挖掘、度量及动态管理机制；其次，构建专家综合影响力驱动的信息集结方法，以及共识度量与反馈调节机制；最后，通过蒙特卡罗模拟获取专家风险系数的最大/最小阈值以及群体风险水平阈值，论证所提模型的合理性与有效性，并将其应用于汽车企业与激光显示科技公司的合作研究中，以辅助实际决策，丰富了社会网络群体共识决策研究。

关键词: 社会网络分析, 群体决策, 信任风险行为, 蒙特卡罗模拟

Abstract:

With the development of social network technology， trust relationships among experts bring new opportunities and challenges to the consensus research of group decision making. Due to the inherent risk of trust， experts may be blindly trusted or lack due trust in decision-making process， which will cause risk issues affecting decision and make it more difficult to reach a scientific and reliable consensus. Therefore， based on experts’ behavior mining， the trust risk in social network group decision making is linked with the scientificity and reliability of the decision result， and a consensus model is proposed for social network group decision making considering experts’ trust risk behaviors mining and dynamic management to reduce the decision risk， improve the credibility of decision result and consensus efficiency. Firstly， based on social network and experts’ credibility， the trust risk behaviors mining， measurement and dynamic management mechanism is designed. Secondly， the information aggregation driven by the comprehensive influence of experts， consensus measure and feedback adjustment mechanism are constructed. Finally， the maximum/minimum thresholds of experts’ risk index and the threshold of group risk level are obtained by Monte Carlo simulation， the rationality and validity of the proposed model are demonstrated， and it is applied to the cooperative research between automobile enterprises and laser display technology companies to assist actual decision-making and enrich the consensus research of social network group decision making.

Key words: social network analysis, group decision making, trust risk behaviors, Monte Carlo simulation

中图分类号:

C934

李梦琪, 刘高峰, 许叶军, 刘霞, 王慧敏. 考虑专家信任风险行为挖掘与动态管理的社会网络群体共识决策研究[J]. 中国管理科学, 2024, 32(11): 144-156.

Mengqi Li, Gaofeng Liu, Yejun Xu, Xia Liu, Huimin Wang. Research on Consensus of Social Network Group Decision Making Considering Experts' Trust Risk Behaviors Mining and Dynamic Management[J]. Chinese Journal of Management Science, 2024, 32(11): 144-156.

图/表 10

图1

表1

表2

表3

图2

表4

表5

图3

表6

具有风险行为的专家动态管理过程"

专家	修正后的偏好信息	修正后的TD
e ₁	$0.5 0.5 0.4 0.15 0.5 0.5 0.6 0.45 0.6 0.4 0.5 0.25 0.85 0.55 0.75 0.5$	0.52
e ₂	$0.5 0.35 0.3 0.4 0.65 0.5 0.4 0.25 0.7 0.6 0.5 0.35 0.6 0.75 0.65 0.5$	/
e ₃	$0.5 0.5 0.25 0.25 0.5 0.5 0.6 0.45 0.75 0.4 0.5 0.2 0.75 0.55 0.8 0.5$	/

表6

表7

算例对比分析结果"

方法	θ_G	t	GRI	专家最终分类	QGDD值	排序结果
本文方法	0.0846	1	0.5808	第一类：e ₁，e ₂，e ₄ 第二类：e ₃，e ₅	QGDD(x ₁)=0.3228，QGDD(x ₂)=0.5252，QGDD(x ₃)=0.4323，QGDD(x ₄)=0.7378	x ₄ $≻$ x ₂ $≻$ x ₃ $≻$ x ₁
模型2	0.1117	1	0.6722	第一类：e ₂，e ₄ 第二类：e ₅ 第三类：e ₁，e ₃	QGDD(x ₁)=0.3283，QGDD(x ₂)=0.5249，QGDD(x ₃)=0.4357，QGDD(x ₄)=0.7221	x ₄ $≻$ x ₂ $≻$ x ₃ $≻$ x ₁
模型3	0.1116	3	0.6763	第一类：e ₄ 第二类：e ₂，e ₅ 第三类：e ₁，e ₃	QGDD(x ₁)=0.4214，QGDD(x ₂)=0.5340，QGDD(x ₃)=0.4565，QGDD(x ₄)=0.5816	x ₄ $≻$ x ₂ $≻$ x ₃ $≻$ x ₁
文献[12]方法	0.1095	4	0.6763	第一类：e ₄ 第二类：e ₂，e ₅ 第三类：e ₁，e ₃	QGDD(x ₁)=0.4309，QGDD(x ₂)=0.5459，QGDD(x ₃)=0.4498，QGDD(x ₄)=0.5674	x ₄ $≻$ x ₂ $≻$ x ₃ $≻$ x ₁

表7

参考文献 25

1	金飞飞，刘金培，陈华友，等. 基于信任关系和信息测度的概率语义社会网络群决策模型［J］. 中国管理科学， 2021， 29（10）： 178-190.
	Jin F F， Liu J P， Chen H Y， et al. Social network group decision-making model based on trust relationship and information measures with probabilistic linguistic information［J］. Chinese Journal of Management Science， 2021， 29（10）： 178-190.
2	Scott H P. Social network analysis： A handbook［M］. London： Sage Publications Ltd.， 2000.
3	Wasserman S， Faust K. Social network analysis： Methods and applications［M］.Cambridge： Cambridge University， 1994.
4	Dong Y C， Zha Q B， Zhang H J， et al. Consensus reaching in social network group decision making： Research paradigms and challenges［J］. Knowledge-Based Systems， 2018， 162（15）： 3-13.
5	Liu X， Xu Y J， Montes R， et al. Social network group decision making： Managing self-confidence-based consensus model with the dynamic importance degree of experts and trust-based feedback mechanism［J］. Information Sciences， 2019， 505： 215-232.
6	Wu J， Cao M S， Chiclana F， et al. An optimal feedback model to prevent manipulation behavior in consensus under social network group decision making［J］. IEEE Transactions on Fuzzy Systems， 2021， 29（7）： 1750-1763.
7	徐选华，侯宇舟. 大群体风险决策理论与方法研究现状及发展趋势［J］. 电子科技大学学报（社科版）， 2021， 23（4）： 68-74.
	Xu X H， Hou Y Z. Research status and development trend of large group risk decision-making theory and method［J］. Journal of University of Electronic Science and Technology of China（Social Sciences Edition）， 2021， 23（4）： 68-74.
8	Xu J， Fung C. A risk-defined trust transitivity model for group decisions in social networks［C］//2019 IFIP/IEEE Symposium on Integrated Network and Service Management （IM）， Arlington， VA， USA，April 8-12， 2019.
9	徐选华，刘尚龙. 社会网络环境下基于“信任—知识”模型的风险性大群体应急决策方法［J］. 运筹与管理. 2021， 30（2）： 31-38.
	Xu X H， Liu S L. Large group risky emergency decision-making method based on trust-knowledge model in social network environment［J］. Operations Research and Management Science， 2021， 30（2）： 31-38.
10	Xu X H， Zhang Q H， Chen X H. Consensus-based non-cooperative behaviors management in large-group emergency decision-making considering experts' trust relations and preference risks［J］. Knowledge-Based Systems， 2020， 190： 105108.
11	徐选华，张前辉. 社会网络环境下保护少数意见的风险性大群体应急决策方法［J］. 运筹与管理，2020， 29（10）： 49-58.
	Xu X H， Zhang Q H. Emergency decision-making method for risky large group protecting minority opinions in social network environment［J］. Operations Research and Management Science，2020，29（10）： 49-58.
12	Li M Q， Xu Y J， Liu X， et al. A trust risk dynamic management mechanism based on third-party monitoring for the conflict-eliminating process of social network group decision making［J］. IEEE Transactions on Cybernetics，2022， 53（6）： 3399-3413.
13	Orlovsky S A. Decision-making with a fuzzy preference relation［J］. Fuzzy Sets and Systems， 1978， 1（3）： 155-167.
14	Xu Y J， Li M Q， Cabrerizo F J， et al. Algorithms to detect and rectify multiplicative and ordinal inconsistencies of fuzzy preference relations［J］. IEEE Transactions on Systems， Man， and Cybernetics： Systems， 2021， 51（6）： 3498-3511.
15	Xia M M， Xu Z S， Chen J. Algorithms for improving consistency or consensus of reciprocal ［0，1］-valued preference relations［J］. Fuzzy Sets and Systems， 2013， 216： 108-133.
16	Chiclana F， Herrera F， Herrera-Viedma E. Integrating three representation models in fuzzy multipurpose decision making based on fuzzy preference relations［J］. Fuzzy Sets and Systems， 1998， 97（1）： 33-48.
17	Pérez Luis G， Mata F， Chiclana F. Social network decision making with linguistic trustworthiness-based induced OWA operators［J］. International Journal of Intelligent Systems， 2014， 29（12）： 1117-1137.
18	Kamis N H， Chiclana F， Levesley J. Preference similarity network structural equivalence clustering based consensus group decision making model［J］. Applied Soft Computing， 2018， 67： 706-720.
19	Victor P， Cornelis C， Cock M D， et al. Practical aggregation operators for gradual trust and distrust［J］. Fuzzy Sets and Systems， 2011， 184（1）： 126-147.
20	Yager R R. On ordered weighted averaging aggregation operators in multicriteria decision making［J］. IEEE Transactions on Systems， Man and Cybernetics， 1988， 18： 183-190.
21	Ureña R， Chiclana F， Melancon G， et al. A social network based approach for consensus achievement in multiperson decision making［J］. Information Fusion， 2019， 47： 72-87.
22	Yager R R. Induced aggregation operators［J］. Fuzzy Sets and Systems， 2003， 137（1）： 59-69.
23	Yager R R. Quantifier guided aggregation using OWA operators［J］. International Journal of Intelligent Systems， 1996， 11（1）： 49-73.
24	Xu Y J， Cabrerizo F J， Herrera-Viedma E. A consensus model for hesitant fuzzy preference relations and its application in water allocation management［J］. Applied Soft Computing， 2017， 58： 265-284.
25	徐选华，马志鹏，陈晓红. 基于公众偏好大数据分析的大群体应急决策质量动态演化研究［J］. 中国管理科学， 2022， 30（7）： 140-149.
	Xu X H， Ma Z P， Chen X H. Dynamic evolution research on emergency decision quality of large group based on the public preferences big data［J］. Chinese Journal of Management Science，2022，30（7）： 140-149.

m n	3	4	5	6	7	8	9
2	0.6211	0.6616	0.6977	0.7083	0.7188	0.7084	0.7095
3	0.6020	0.6606	0.6868	0.6913	0.6956	0.6982	0.6937
4	0.5840	0.6418	0.6788	0.6819	0.6867	0.6898	0.6916
5	0.5738	0.6314	0.6687	0.6779	0.6802	0.6798	0.6811
6	0.5647	0.6164	0.6590	0.6717	0.6774	0.6764	0.6802
7	0.5590	0.6144	0.6530	0.6678	0.6753	0.6729	0.6763
8	0.5506	0.6060	0.6450	0.6666	0.6675	0.6703	0.6711
9	0.5466	0.6075	0.6451	0.6619	0.6696	0.6660	0.6682
10	0.5371	0.5966	0.6339	0.6599	0.6622	0.6662	0.6661
11	0.5362	0.5929	0.6338	0.6545	0.6640	0.6639	0.6660
12	0.5357	0.5865	0.6282	0.6534	0.6595	0.6655	0.6619
13	0.5254	0.5841	0.6256	0.6487	0.6578	0.6620	0.6611
14	0.5287	0.5823	0.6275	0.6493	0.6553	0.6581	0.6590
15	0.5209	0.5792	0.6266	0.6476	0.6548	0.6581	0.6565
16	0.5200	0.5725	0.6250	0.6464	0.6517	0.6524	0.6568
17	0.5109	0.5730	0.6222	0.6441	0.6488	0.6539	0.6559
18	0.5097	0.5709	0.6175	0.6441	0.6489	0.6511	0.6577
19	0.5044	0.5630	0.6179	0.6397	0.6471	0.6501	0.6496

m n	3	4	5	6	7	8	9
2	0.6980	0.7282	0.7550	0.7607	0.7698	0.7582	0.7605
3	0.7158	0.7542	0.7720	0.7705	0.7738	0.7743	0.7718
4	0.7238	0.7605	0.7790	0.7780	0.7808	0.7836	0.7835
5	0.7268	0.7611	0.7823	0.7870	0.7851	0.7861	0.7852
6	0.7298	0.7593	0.7830	0.7866	0.7937	0.7891	0.7916
7	0.7330	0.7625	0.7833	0.7918	0.7952	0.7913	0.7939
8	0.7332	0.7635	0.7824	0.7933	0.7916	0.7940	0.7939
9	0.7339	0.7676	0.7854	0.7930	0.7982	0.7951	0.7970
10	0.7319	0.7648	0.7818	0.7955	0.7954	0.7992	0.7977
11	0.7342	0.7639	0.7847	0.7955	0.8003	0.7988	0.8011
12	0.7364	0.7631	0.7836	0.7954	0.7985	0.8023	0.7991
13	0.7331	0.7624	0.7833	0.7950	0.7998	0.8022	0.8015
14	0.7387	0.7650	0.7879	0.7977	0.7995	0.8009	0.8017
15	0.7338	0.7644	0.7875	0.7979	0.8015	0.8026	0.8023
16	0.7351	0.7604	0.7888	0.7989	0.8020	0.8007	0.8035
17	0.7320	0.7630	0.7876	0.7990	0.8011	0.8034	0.8051
18	0.7315	0.7634	0.7857	0.8004	0.8018	0.8025	0.8068
19	0.7303	0.7600	0.7877	0.7982	0.8016	0.8033	0.8033

m n	3	4	5	6	7	8	9
2	0.6595	0.6949	0.7264	0.7345	0.7443	0.7333	0.7350
3	0.6645	0.7121	0.7326	0.7346	0.7373	0.7389	0.7360
4	0.6674	0.7108	0.7365	0.7359	0.7395	0.7417	0.7400
5	0.6658	0.7128	0.7374	0.7417	0.7411	0.7393	0.7385
6	0.6675	0.7111	0.7386	0.7395	0.7435	0.7388	0.7410
7	0.6712	0.7124	0.7355	0.7439	0.7449	0.7377	0.7401
8	0.6708	0.7146	0.7348	0.7444	0.7402	0.7403	0.7375
9	0.6708	0.7168	0.7377	0.7447	0.7420	0.7383	0.7393
10	0.6676	0.7133	0.7332	0.7429	0.7428	0.7400	0.7388
11	0.6708	0.7164	0.7366	0.7419	0.7423	0.7404	0.7387
12	0.6735	0.7128	0.7373	0.7417	0.7428	0.7412	0.7395
13	0.6714	0.7124	0.7352	0.7409	0.7414	0.7429	0.7384
14	0.6747	0.7154	0.7386	0.7429	0.7418	0.7396	0.7378
15	0.6699	0.7167	0.7375	0.7424	0.7416	0.7413	0.7375
16	0.6716	0.7135	0.7397	0.7432	0.7419	0.7393	0.7384
17	0.6701	0.7160	0.7375	0.7446	0.7418	0.7399	0.7385
18	0.6707	0.7154	0.7364	0.7433	0.7411	0.7402	0.7394
19	0.6696	0.7158	0.7379	0.7433	0.7419	0.7399	0.7361

模型	n	m	θ_G 均值	t均值	cs均值	GRI均值
模型1	4	5	0.1476	0.7210	0.9830	0.6422
		11	0.1794	0.8730	0.9930	0.6568
		17	0.1889	1.5250	0.9320	0.6575
	7	5	0.1580	0.6660	0.9790	0.6864
		11	0.1892	1.2350	0.9910	0.6971
		17	0.1969	3.0670	0.6960	0.7015
模型2	4	5	0.1553	0.1450	1	0.7027
		11	0.1843	0.7720	0.9780	0.7043
		17	0.1925	1.8010	0.8520	0.7011
	7	5	0.1666	0.0410	1	0.7395
		11	0.1927	1.2830	0.9760	0.7325
		17	0.1998	3.6520	0.5150	0.7341
模型3	4	5	0.1800	0.9570	0.9980	0.6821
		11	0.1936	2.9340	0.8250	0.6882
		17	0.2009	4.1790	0.4160	0.6912
	7	5	0.1864	0.9010	1	0.7229
		11	0.1946	3.7150	0.7100	0.7347
		17	0.2053	4.8580	0.1030	0.7380
模型4	4	5	0.1898	1.8430	0.6390	0.6725
		11	0.2207	1.9520	0.1290	0.6924
		17	0.2279	2.1310	0.0340	0.6982
	7	5	0.1927	1.9650	0.6500	0.7145
		11	0.2203	2.4010	0.0650	0.7278
		17	0.2291	2.7060	0.001	0.7330

研究类型	考虑信任风险的社会网络群体共识决策研究				基于信任关系的社会网络群体共识决策研究
性能方法	本文方法	[12]	[11]	[10]	[1, 4-6, 18, 19]
信任传递与集结	✕	✕	✕	✕	􀳫
专家决策行为（非合作行为等）	􀳫	􀳫	􀳫	􀳫	􀳫
专家信任风险行为挖掘及管理	􀳫	􀳫	✕	✕	✕
促进共识达成的能力	􀳫	-	􀳫	􀳫	􀳫
可接受的决策成本	􀳫	-	􀳫	􀳫	􀳫
控制群体风险水平的能力	􀳫	-	✕	✕	✕