基于多项式主观逻辑的扩展信任传播模型

doi:10.3969/j.issn.1000-436x.2013.05.002

摘要/Abstract

摘要：

摘要：为了提高信任评价的准确性和主观逻辑的适用性，提出了基于多项式主观逻辑的扩展信任传播模型，该模型由信任融合和信任传递2部分组成。基于信誉的多项式观点的融合方法考虑信誉与观察环境对融合操作的影响，提高了信任评价的准确性，更符合人的直觉评判；信任传递把多项式观点的传递理解为传递过程是二项逻辑观点，而传递信息是多项式观点，给出了多项式观点的信任传递公式，并对相关定理进行了证明。最后，通过实例分析验证了模型的准确性和适应性。

关键词: 主观逻辑扩展, 多项式观点, 信任传播, 信任融合, 信任传递

Abstract:

In order to improve the accurateness of trust evaluation and the applicability of subjective logic,a multinomial subjective logic based extended trust propagation model (MSL-ETPM) was presented,which was composed of trust fu-sion and trust transitivity.Making the fused opinion more accurate and more in line with the intuitive judgment of human,reputation-based fusion operator of multinomial opinions takes effects of the reputation value and observation circum-stance into consideration.Trust transitivity interprets multinomial opinion transitivity as meaning that transitivity process is binomial logical opinion and transitivity information is multinomial opinion,giving trust transitivity formula of multi-nomial opinion and proving related theorem.Finally,the accuracy and adaptability of the model are put to the proof by example analysis.

Key words: extended subjective logic, multinomial opinion, trust propagation, trust fusion, trust transitivity

田俊峰,吴丽娟. 基于多项式主观逻辑的扩展信任传播模型[J]. 通信学报, 2013, 34(5): 12-19.

Jun-feng TIAN,Li-juan WU. Multinomial subjective logic based extended trust propagation model[J]. Journal on Communications, 2013, 34(5): 12-19.

图/表 5

图1

图2

表1

表2

表3

在不同情况下Zadeh例子应用不同融合操作所得结果的比较"

模型	实体关系	依赖类型	依赖度/信誉值	Peter	Paul	Mary	Θ	序号
J?sang		观点独立观点依赖	η_t1 2=0,η_t2 1=0η_t1 2=1,η_t2 1=1	0.492 50.49	0.010 00.01	0.492 50.49	0.005 00.01	12
J?sang		部分依赖	η_tw1w2=0.8,η_tw2w1=0.4η_tw1w2=0.4,η_tw2w1=0.8	0.421 20.561 6	0.010 00.010 0	0.561 60.421 2	0.007 20.007 2	34
	信誉相同，环境相同rep^w1=rep^w2=1,η_ew1w2=η_ew2w1=1	观点独立观点依赖	η_tw1w2=0,η_tw2w1=0η_tw1w2=1,η_tw2w1=1	0.493 70.492 5	0.010 10.010 0	0.493 70.492 5	0.002 50.005 0	56
	部分依赖	η_tw1w2=0.8,η_tw2w1=0.4η_tw1w2=0.4,η_tw2w1=0.8	0.422 70.563 6	0.010 10.010 1	0.563 60.422 7	0.003 60.003 6	78
	信誉相同，环境不同rep^w1=rep^w2=1,η_ew1w2=1,η_ew2w1=0.3	观点独立观点依赖	η_tw1w2=0,η_tw2w1=0η_tw1w2=1,η_tw2w1=1	0.441 90.419 4	0.010 10.010 1	0.545 00.566 2	0.003 00.004 3	910
	部分依赖	η_tw1w2=0.8,η_tw2w1=0.4η_tw1w2=0.4,η_tw2w1=0.8	0.394 50.466 2	0.010 10.010 1	0.591 80.520 0	0.003 60.003 7	1112
		观点独立	rep^w1=0.45,rep^w2=0.65	0.402 5	0.010 0	0.581 4	0.006 1	13
		η_tw1w2=0,η_tw2w1=0	rep^w1=0.2,rep^w2=0.9	0.178 9	0.010 0	0.805 0	0.006 1	14
		观点依赖	rep^w1=0.45,rep^w2=0.65	0.401 6	0.010 0	0.579 6	0.009 1	15
MSL-ETPM	信誉不同，环境相同η_ew1w2=η_ew2w1=1	η_tw1w2=1,η_tw2w1=1	rep^w1=0.2,rep^w2=0.9	0.178 3	0.010 0	0.802 6	0.009 1	16
	部分依赖	rep^w1=0.45,rep^w2=0.65η_tw1w2=0.8,η_tw2w1=0.4	0.362 9	0.010 0	0.619 6	0.007 5	17
	rep^w1=0.2,rep^w2=0.9η_tw1w2=0.8,η_tw2w1=0.4	0.155 6	0.010 0	0.827 2	0.007 2	18
	rep1=0.45,rep2=0.65η_tw1w2=0.4,η_tw2w1=0.8	0.442 0	0.010 0	0.540 3	0.007 7	19
	rep^w1=0.2,rep^w2=0.9η_tw1w2=0.4,η_tw2w1=0.8	0.204 3	0.010 0	0.777 7	0.008 0	20
		观点独立	rep^w1=0.45,rep^w2=0.65	0.354 0	0.010 0	0.630 6	0.005 4	21
		η_tw1w2=0,η_tw2w1=0	rep^w1=0.2,rep^w2=0.9	0.150 4	0.010 0	0.834 5	0.005 1	22
		观点依赖	rep^w1=0.45,rep^w2=0.65	0.333 0	0.010 0	0.649 4	0.007 6	23
	信誉不同，环境不同η_ew1w2=1,η_ew2w1=0.3	η_tw1w2=1,η_tw2w1=1	rep^w1=0.2,rep^w2=0.9	0.138 9	0.010 0	0.844 0	0.007 1	24
		部分依赖	rep^w1=0.45,rep^w2=0.65η_tw1w2=0.8,η_tw2w1=0.4	0.310 6	0.010 0	0.673 0	0.006 4	25
		rep^w1=0.2,rep^w2=0.9η_tw1w2=0.8,η_tw2w1=0.4	0.127 0	0.010 0	0.857 1	0.005 9	26
		rep^w1=0.45,rep^w2=0.65η_tw1w2=0.4,η_tw2w1=0.8	0.376 6	0.010 0	0.606 8	0.006 6	27
		rep^w1=0.2,rep^w2=0.9η_tw1w2=0.4,η_tw2w1=0.8	0.127 0	0.010 0	0.857 1	0.005 9	28

表3

参考文献 24

[1]	唐文，陈钟 . 基于模糊集合理论的主观信任管理模型研究[J]. 软件学报， 2003,14(8):1401-1408.
[2]	田俊峰，杜瑞忠，刘玉玲 . 基于结点行为特征的可信性度量模型[J]. 计算机研究与发展， 2011,48(6):934-944.
[3]	AUDUN J , STEPHEN M , SIMON P . Exploring different types of trust propagation[A].Proceedings of the 4th International Conference on Trust Management(iTrust)[C]. Pisa,Italy, 2006. 179-192.
[4]	BLAZE M , FEIGENBAUM J . Decentralized trust management[A].Proceedings of the 17th Symposium on Security and Privacy[C]. Okaland,USA, 1996. 164-173.
[5]	THOMAS B , MALTE B , BIRGIT K . Valuation of trust in open net-works.GOLLMANN D[A].Proceedings of the 3rd European Sympo-sium on Research in Computer Security[C]. Brighton: Springer-Veriag, 1994. 3-18.
[6]	ALFAREZ A STEPHEN H . A distributed trust model[A]. Proceedings of the 1997 Workshop on New Security Paradigms[C]. 1997. 48-60.
[7]	WANG Y , VASSILEVA J . Bayesian network-based trust model[A].Proceedings of the IEEE Computer Society WIC International Con-ference on Web Intelligence[C]. 2003. 372-378.
[8]	GUHA R , KUMAR R , RAGHAVAN P , et al . Propagation of trust and distrust[A].Proceedings of the 13th International Conference on World Wide Web[C]. 2004. 17-22.
[9]	YU B , SINGH P . An evidential model of distributed reputation man-agement[A].Proceedings of the 1st international joint conference on autonomous agents and multi agent systems: part 1[C]. 2002. 294-301.
[10]	YU B , SINGH P . Searching social networks[A].Proceedings of the Second Intl Joint Conference on Autonomous Agents ＆ Multiagent Systems[C]. 2003.
[11]	AUDUN J . A subjective metric of authentication[A].Proceedings of the 5th European Symposium on Research in Computer Security[C]. 1998. 329-444
[12]	张喜征 . 虚拟企业信任机制研究—网络环境下信任管理模式创新[M]. 长沙：湖南人民出版社， 2005.
[13]	AUDUN J . Fission of opinions in subjective logic[A].Proceedings of the 12th International Conference on Information Fusion (FUSION 2009)[C]. Seattle,USA, 2009.
[14]	AUDUN J , O'Hara S . Multiplication of multinomial subjective opin-ions[A].Proceedings of the 13th International Conference on Informa-tion Processing and Management of Uncertainty (IPMU2010)[C]. Dortmund,Germany, 2010. 248-257.
[15]	AUDUN J . A logic for uncertain probabilities[J]. International Journal of Uncertainty,Fuzziness and Knowledge-Based Systems, 2001,9(3):1-31.
[16]	AUDUN J , ZIED E . Interpreting belief functions as dirichlet distribu-tions[A].Proceedings of the 9th European Conference on Symbolic and Quantitative Approaches to Reasoning with Uncertainty (EC-SQARU)[C]. Hammamet,Tunisia, 2007. 393-404.
[17]	AUDUN J , JAVIER D , MARIA R . Cumulative and averaging fusion of beliefs[J]. Information Fusion, 2010,11(2):192-200.
[18]	AUDUN J . The consensus operator for combining beliefs[J]. Artificial Intelligence Journal, 2002,142((1-2)):157-170.
[19]	AUDUN J , TOUHID B , XU Y , et al . Combining trust and reputation management for Web-based services[A].Proceedings of the 5th Inter-national Conference on Trust,Privacy ＆ Security in Digital Busi-ness(TrustBus2008)[C]. Turin,Italy, 2008. 90-99.
[20]	JфSANG A , POPE S . Semantic constraints for trust transitivity[A].Proc Second Asia-Pacific Conference on Conceptual Modeling (APCCM2005)[C]. Neweastle,Australia, 2005. 59-68.
[21]	LOTFI Z . Review of Shafer's a mathematical theory of evidence[J]. AI Magazine, 1984(5):81-83.
[22]	SMETS P . Belief functions[A].Non Standard Logics for Automated Reason-ing[C]. London,UK, 1988. 253-286.
[23]	FESTINGER L . A Theory of Cognitive Dissonance[M]. Stanford: Stanford University Press, 1957.
[24]	ARONSON E . Back to the future: retrospective review of Leon Festinger's: a theory of cognitive dissonance[J].The American Journal of Psychology, 1997,110(1):127-137.

规则	Peter	Paul	Mary	Θ
Dempster's 规则	0.490	0.015	0.490	0.005
非正规化规则	0.009 8	0.000 3	0.009 8	0.000 1