基于Hopfield神经网络的谣言认知模型研究
|
摘要
谣言是一种特殊的社会舆论现象,很容易引起人们的恐慌行为,对社会正常的生产和生活秩序有着很不利的影响。随着日新月异的互联网和无线通讯网的高速发展,谣言的传播范围更为广泛,因此对谣言内在传播机制的探索和研究就变得更为复杂和困难。目前,国内外的学者分别从社会心理学和信息科学角度对一般的社会谣言进行了研究,取得了一定的进展和成果,但效果并不十分明显,对谣言传播和蔓延的规律以及谣言产生的本质特征及内在机理仍未给出一个准确和可靠的结论。
地震谣言作为一种特殊的社会谣言,具有一般社会谣言的基本特点,但由于其特殊的背景,成为社会公众中关注度高、影响力大的一种社会谣言,因此本文拟以地震谣言为背景来对谣言的产生机制和传播规律做一些研究工作。通过对一系列地震谣言事件的观察发现,个体差异性在谣言的形成及传播过程中表现得十分明显,不同的个体本身具有不同的特征属性,因此对同一信息的认知会有所差异,由此可见对个体的认知模型进行研究对研究个体的谣言传播行为是非常有意义的。Hopfield神经网络具有联想记忆的功能,而且相关生物学研究表明人脑负责联想记忆功能的海马CA3区中的突触连接和Hopfield网络神经元的连接具有相似的连接特性,因此本文采用Hopfield神经网络来对个体的认知模型进行了建模,通过Hopfield神经网络的联想记忆功能来模拟人脑的海马CA3区的联想记忆功能,从而实现个体对谣言信息的记忆和识别,最后本文通过C++编程对个体的认知模型进行了实现并进行了一系列的实验和数据分析。在实验阶段,本文通过让不同的个体学习不同的样本信息来表现个体本身的差异性,然后再观察不同个体对同一信息的识别结果有何不同,并与现实情况进行了对比分析。通过对实验数据的分析可以发现,本文所建立的个体认知模型与现实个体的认知基本相符。
As a general public phenomenon, rumors usually disseminate in a face-to-face and word-of-mouth way, whose communication effect and range are somewhat limited. However, with the rapid development of the Internet and wireless networks, the rumors spread to wider range recent years, so the exploration and study to the rumors’inner spread mechanism become more complex and difficult. At present, the domestic and foreign scholars have done some researches to the general social rumors respectively in terms of social psychology and information science, and have made some progress and achievements, but the effect is not very obvious. An accurate and reliable conclusion is still not given out about the law of the rumors’spread and the essential characteristics and internal mechanism of rumors’engenderment.
Earthquake rumors, as a special kind of social rumors, have the basic characteristics of general society rumors, but due to its special background, it attracts high attention and has a great influence in the social public. So this paper intends to do some research work to the generation mechanism and spread rule of the rumors at the background of earthquake rumors. Through observation to a series of events that of earthquake rumors, individual differences in the process of the formation and spreading of rumors fared obviously. Different individual has different characteristic attributes, so the cognition to the same information will vary, thus to study the individual cognitive model is very significant to the individual rumor spread behavior. Hopfield neural network has an associative memory function, and the biology research shows that the synapses’junctions in the hippocampus CA3 area, the part of brain responsible for associative memory function have similar characteristics with the neurons in the Hopfield network, and hence Hopfield neural network is adopted in this paper to individual cognitive model, through the associative memory function of Hopfield neural network to simulate the associative memory function of the human brain CA3, so as to realize the individual’s memory and identification to information of rumors. Finally, this paper realizes the individual cognitive model through c + + programming and conducts a series of experiments and data analysis, and gives a comparative analysis with the reality. Based on the It can be found, through analysis to the experimental data, individual cognitive model set up in the paper basic conform to the real individual cognition.
地震谣言作为一种特殊的社会谣言,具有一般社会谣言的基本特点,但由于其特殊的背景,成为社会公众中关注度高、影响力大的一种社会谣言,因此本文拟以地震谣言为背景来对谣言的产生机制和传播规律做一些研究工作。通过对一系列地震谣言事件的观察发现,个体差异性在谣言的形成及传播过程中表现得十分明显,不同的个体本身具有不同的特征属性,因此对同一信息的认知会有所差异,由此可见对个体的认知模型进行研究对研究个体的谣言传播行为是非常有意义的。Hopfield神经网络具有联想记忆的功能,而且相关生物学研究表明人脑负责联想记忆功能的海马CA3区中的突触连接和Hopfield网络神经元的连接具有相似的连接特性,因此本文采用Hopfield神经网络来对个体的认知模型进行了建模,通过Hopfield神经网络的联想记忆功能来模拟人脑的海马CA3区的联想记忆功能,从而实现个体对谣言信息的记忆和识别,最后本文通过C++编程对个体的认知模型进行了实现并进行了一系列的实验和数据分析。在实验阶段,本文通过让不同的个体学习不同的样本信息来表现个体本身的差异性,然后再观察不同个体对同一信息的识别结果有何不同,并与现实情况进行了对比分析。通过对实验数据的分析可以发现,本文所建立的个体认知模型与现实个体的认知基本相符。
As a general public phenomenon, rumors usually disseminate in a face-to-face and word-of-mouth way, whose communication effect and range are somewhat limited. However, with the rapid development of the Internet and wireless networks, the rumors spread to wider range recent years, so the exploration and study to the rumors’inner spread mechanism become more complex and difficult. At present, the domestic and foreign scholars have done some researches to the general social rumors respectively in terms of social psychology and information science, and have made some progress and achievements, but the effect is not very obvious. An accurate and reliable conclusion is still not given out about the law of the rumors’spread and the essential characteristics and internal mechanism of rumors’engenderment.
Earthquake rumors, as a special kind of social rumors, have the basic characteristics of general society rumors, but due to its special background, it attracts high attention and has a great influence in the social public. So this paper intends to do some research work to the generation mechanism and spread rule of the rumors at the background of earthquake rumors. Through observation to a series of events that of earthquake rumors, individual differences in the process of the formation and spreading of rumors fared obviously. Different individual has different characteristic attributes, so the cognition to the same information will vary, thus to study the individual cognitive model is very significant to the individual rumor spread behavior. Hopfield neural network has an associative memory function, and the biology research shows that the synapses’junctions in the hippocampus CA3 area, the part of brain responsible for associative memory function have similar characteristics with the neurons in the Hopfield network, and hence Hopfield neural network is adopted in this paper to individual cognitive model, through the associative memory function of Hopfield neural network to simulate the associative memory function of the human brain CA3, so as to realize the individual’s memory and identification to information of rumors. Finally, this paper realizes the individual cognitive model through c + + programming and conducts a series of experiments and data analysis, and gives a comparative analysis with the reality. Based on the It can be found, through analysis to the experimental data, individual cognitive model set up in the paper basic conform to the real individual cognition.
引文
[1]刘仲达.新传播手段下的地震谣传和对策思考[J].学理论,2009,(28):141- 142.
[2]奥尔波特等.谣言心理学[M].沈阳:辽宁教育出版社,2003.
[3]汪志坚,骆少康.网络上流传谣言类型与特性研究[D].2002年网络与社会研讨会论文(台湾).
[4]胡钰.大众传播效果[M].北京:新华出版社, 2000.
[5]巢乃鹏,黄娴.网络传播中的“谣言”现象研究[J].情报理论与实践.2004,27(6):586-589,575.
[6] Serge Galam. Modeling Rumors - the no Plane Pentagon French Hoax Case [J]. Physica A, 2003 320:571-580.
[7] M. Nekovee, Y. Morenob, etl. Theory of Rumor Spreading in Complex Social Networks [J]. Physica A, 2007, 374: 457-470.
[8] William Basener, Bernard P. Brooks,etl. The Brouwer Fixed Point Theorem Applied to Rumour Transmission [J]. Applied Mathematics Letters, 2006, 19: 841–842.
[9] Kazuki Kawachi. Deterministic Models for Rumor Transmission [J]. Nonlinear Analysis: Real World Applications, 2008, 9:1989-2028.
[10] Kazuki Kawachi, Motohide Seki, et al. A Rumor Transmission Model with Various Contact Interactions [J]. Journal of Theoretical Biology. 2008,253:55-60.
[11] D.J.Watts, S.H.Strogatz. Collective dynamics of’small-world’Networks [J]. Nature, 1998, 393: 440 -442.
[12]黄金煜.小世界网络上的谣言免疫策略研究[D].杭州:浙江大学,2008.
[13]刘常昱,胡晓峰,司光亚,罗批.基于小世界网络的舆论传播模型研究[J].系统仿真学报. 2006, 18(12):3608-3610.
[14]刘常昱,胡晓峰,罗批,司光亚.基于不对称人际影响的舆论涌现模型研究[J].系统仿真学报.2008, 20(4):990-996.
[15]刘常昱,胡晓峰,司光亚,罗批.舆论涌现模型研究[J].复杂系统与复杂性科学,2007,4(1):22-27.
[16]张文修,徐伟华.基于粒计算的认知模型[J].工程数学学报,2007,6(24):957-971.
[17] NECHES R, FIKES R E, FININ T, et a1.Enabling technology for knowledge sharing [J].AI magazine,1991,12(3):36-56.
[18] GRUBER T R. A translation approach to portable ontology specifications [J]. Knowledge Acquistion, 1993, 5(2):199-200.
[19] BORST W N. Construction of Engineering Ontologies for Knowledge Sharingand Reuse [D]. Enschede: University of Twente, 1997.
[20] STUDER R, BENJAMlNS V R, FENSEL D. Knowledge Engineering, Principles and Methods [J]. Data and Knowledge Engineering, 1998, 25(122): 161-197.
[21] CHANDRASEKARAN B,JOSEPHSON J R, BENJAMINS V R. What Are Ontologies and Why Do We Need Them? [J]. IEEE Intelligent Systems, 1999, 14(1):20-26.
[22] WANG Y. Novel Approaches in Cognitive informatics and Natural Intelligence [M]. USA: IGI Global, 2008.
[23] WANG Y, SHEU P. Toward Cognitive Computers that Learn and Think [J]. ACM Transactions on Autonomous and Adaptive Systems, 2008, 2(4).
[24] WANG Y. Cognitive Informatics:Exploring Theoretical Foundations for Natural Intelligence, Neural Informatics, Autonomic Computing, and Agent Systems [J].The international Journal of Cognitive Informatics and Natural Intelligence(IJCiNi),2007,l(1):1-10.
[25] WANG Y. Cognitive informatics Models of the Brain[J]. IEEE Transactions on Systems, Man, and Cybernetics(C),2006,36(2):203-207.
[26] WANG Y. On Cognitive Informatics [J]. Brain and Mind: A Transdisciplinary Journal of Neuroscience and Neurophilosophy,2003,4(2):151—167.
[27] MIO J S, STERNBERG R J. Cognitive Psychology [M]. Belmont, Calif.: Wadsworth Pub Co inc, 2005.
[28]孙振凯,苗崇刚,顾建华,邹其嘉,杨懋源.近年来我国地震谣言的起因及对策[J].国际地震动态,2001,(1):12 - 17.
[29]刘在涛.近年地震谣言特点与平息对策建议[J].中国应急救援,2010,(4):16-19.
[30]陈宇卫,张鹏.一起特大地震谣传事件后的反思与启示[J].国际地震动态,1998,(9):28 - 30.
[31]赵旺凶,乔清理,王丹.基于类Hopfield脉冲神经网络模型由突触缺失导致海马CA3区联想记忆功能障碍的建模仿真研究[J].国际生物医学工程杂志,2010,33(4):201-204,211.
[32] Bierman EJ, Comijs HC, Jonker C, et al.The effect of anxiety and depression on decline of memory function in Alzheimer's disease [J]. Int Psychogeriatr, 2009,2l(6):1142-1147.
[33] Perus M.Neural networks as a basis for quantum associative networks[J].Neural Netw World,2000,10(6):1001—1013.
[34] Lorente de N6 R. Studies on the structure of the cerebral cortex. II. Continuationof the study of the ammonic system [J]. J Psychol Neurol,1934,46(2):l13-177.
[35] Li XG, Somogyi P, Ylinen A, et a1. The hippocampal CA3 network: an in vivo intracellular labeling study[J].J Comp Neurol, 1994, 339(2):181-208.
[36] Samura T, Hattori M, Ishizaki S. Sequence disambiguation and pattern completion by cooperation between autoassociative and heteroasseeiative memories of functionally divided hippocampal CA3[J]. Neurocomputing, 2008,71(16-18):3176-3183.
[37] De Almeida L, Idiart M, Lisman JE.Memory retrieval time and memory capacity of the CA3 network:role of gamma frequency oscillations[J].Learn Mem,2007,14(11):795-806.
[38] Hasegawa H. An associative memory of Hodgkin-huxley neuron networks with willshaw-type synaptic couplings [J]. J Phys Sec Jpn,2001,70(7):2210-2219.
[39] Horn D, Levy N, Ruppin E. Neuronal-based synaptic compensation: a computational study in Alzheimer's disease[J].Neural Comput,1996,8(6):1227- 1243.
[40] Izhikevich EM. Simple model of spiking neurons [J].IEEE Trans Neural Netw, 2003, 14(6):1569-1572.
[41] Faisal AA, Selen LP, Wolpert DM. Noise in the nervous system[J].Nat Rev Neurosci,2008,9(4):292-303.
[42]朱凯,王正林.精通MATLAB神经网络[M].北京:电子工业出版社,2010.
[43]夏松.Hopfield神经网络构造的联想记忆存储器实现与研究[D].安徽大学,2010,(10).
[44]王剑.一类新型联想记忆神经网络研究[D].华南理工大学,2003.
[2]奥尔波特等.谣言心理学[M].沈阳:辽宁教育出版社,2003.
[3]汪志坚,骆少康.网络上流传谣言类型与特性研究[D].2002年网络与社会研讨会论文(台湾).
[4]胡钰.大众传播效果[M].北京:新华出版社, 2000.
[5]巢乃鹏,黄娴.网络传播中的“谣言”现象研究[J].情报理论与实践.2004,27(6):586-589,575.
[6] Serge Galam. Modeling Rumors - the no Plane Pentagon French Hoax Case [J]. Physica A, 2003 320:571-580.
[7] M. Nekovee, Y. Morenob, etl. Theory of Rumor Spreading in Complex Social Networks [J]. Physica A, 2007, 374: 457-470.
[8] William Basener, Bernard P. Brooks,etl. The Brouwer Fixed Point Theorem Applied to Rumour Transmission [J]. Applied Mathematics Letters, 2006, 19: 841–842.
[9] Kazuki Kawachi. Deterministic Models for Rumor Transmission [J]. Nonlinear Analysis: Real World Applications, 2008, 9:1989-2028.
[10] Kazuki Kawachi, Motohide Seki, et al. A Rumor Transmission Model with Various Contact Interactions [J]. Journal of Theoretical Biology. 2008,253:55-60.
[11] D.J.Watts, S.H.Strogatz. Collective dynamics of’small-world’Networks [J]. Nature, 1998, 393: 440 -442.
[12]黄金煜.小世界网络上的谣言免疫策略研究[D].杭州:浙江大学,2008.
[13]刘常昱,胡晓峰,司光亚,罗批.基于小世界网络的舆论传播模型研究[J].系统仿真学报. 2006, 18(12):3608-3610.
[14]刘常昱,胡晓峰,罗批,司光亚.基于不对称人际影响的舆论涌现模型研究[J].系统仿真学报.2008, 20(4):990-996.
[15]刘常昱,胡晓峰,司光亚,罗批.舆论涌现模型研究[J].复杂系统与复杂性科学,2007,4(1):22-27.
[16]张文修,徐伟华.基于粒计算的认知模型[J].工程数学学报,2007,6(24):957-971.
[17] NECHES R, FIKES R E, FININ T, et a1.Enabling technology for knowledge sharing [J].AI magazine,1991,12(3):36-56.
[18] GRUBER T R. A translation approach to portable ontology specifications [J]. Knowledge Acquistion, 1993, 5(2):199-200.
[19] BORST W N. Construction of Engineering Ontologies for Knowledge Sharingand Reuse [D]. Enschede: University of Twente, 1997.
[20] STUDER R, BENJAMlNS V R, FENSEL D. Knowledge Engineering, Principles and Methods [J]. Data and Knowledge Engineering, 1998, 25(122): 161-197.
[21] CHANDRASEKARAN B,JOSEPHSON J R, BENJAMINS V R. What Are Ontologies and Why Do We Need Them? [J]. IEEE Intelligent Systems, 1999, 14(1):20-26.
[22] WANG Y. Novel Approaches in Cognitive informatics and Natural Intelligence [M]. USA: IGI Global, 2008.
[23] WANG Y, SHEU P. Toward Cognitive Computers that Learn and Think [J]. ACM Transactions on Autonomous and Adaptive Systems, 2008, 2(4).
[24] WANG Y. Cognitive Informatics:Exploring Theoretical Foundations for Natural Intelligence, Neural Informatics, Autonomic Computing, and Agent Systems [J].The international Journal of Cognitive Informatics and Natural Intelligence(IJCiNi),2007,l(1):1-10.
[25] WANG Y. Cognitive informatics Models of the Brain[J]. IEEE Transactions on Systems, Man, and Cybernetics(C),2006,36(2):203-207.
[26] WANG Y. On Cognitive Informatics [J]. Brain and Mind: A Transdisciplinary Journal of Neuroscience and Neurophilosophy,2003,4(2):151—167.
[27] MIO J S, STERNBERG R J. Cognitive Psychology [M]. Belmont, Calif.: Wadsworth Pub Co inc, 2005.
[28]孙振凯,苗崇刚,顾建华,邹其嘉,杨懋源.近年来我国地震谣言的起因及对策[J].国际地震动态,2001,(1):12 - 17.
[29]刘在涛.近年地震谣言特点与平息对策建议[J].中国应急救援,2010,(4):16-19.
[30]陈宇卫,张鹏.一起特大地震谣传事件后的反思与启示[J].国际地震动态,1998,(9):28 - 30.
[31]赵旺凶,乔清理,王丹.基于类Hopfield脉冲神经网络模型由突触缺失导致海马CA3区联想记忆功能障碍的建模仿真研究[J].国际生物医学工程杂志,2010,33(4):201-204,211.
[32] Bierman EJ, Comijs HC, Jonker C, et al.The effect of anxiety and depression on decline of memory function in Alzheimer's disease [J]. Int Psychogeriatr, 2009,2l(6):1142-1147.
[33] Perus M.Neural networks as a basis for quantum associative networks[J].Neural Netw World,2000,10(6):1001—1013.
[34] Lorente de N6 R. Studies on the structure of the cerebral cortex. II. Continuationof the study of the ammonic system [J]. J Psychol Neurol,1934,46(2):l13-177.
[35] Li XG, Somogyi P, Ylinen A, et a1. The hippocampal CA3 network: an in vivo intracellular labeling study[J].J Comp Neurol, 1994, 339(2):181-208.
[36] Samura T, Hattori M, Ishizaki S. Sequence disambiguation and pattern completion by cooperation between autoassociative and heteroasseeiative memories of functionally divided hippocampal CA3[J]. Neurocomputing, 2008,71(16-18):3176-3183.
[37] De Almeida L, Idiart M, Lisman JE.Memory retrieval time and memory capacity of the CA3 network:role of gamma frequency oscillations[J].Learn Mem,2007,14(11):795-806.
[38] Hasegawa H. An associative memory of Hodgkin-huxley neuron networks with willshaw-type synaptic couplings [J]. J Phys Sec Jpn,2001,70(7):2210-2219.
[39] Horn D, Levy N, Ruppin E. Neuronal-based synaptic compensation: a computational study in Alzheimer's disease[J].Neural Comput,1996,8(6):1227- 1243.
[40] Izhikevich EM. Simple model of spiking neurons [J].IEEE Trans Neural Netw, 2003, 14(6):1569-1572.
[41] Faisal AA, Selen LP, Wolpert DM. Noise in the nervous system[J].Nat Rev Neurosci,2008,9(4):292-303.
[42]朱凯,王正林.精通MATLAB神经网络[M].北京:电子工业出版社,2010.
[43]夏松.Hopfield神经网络构造的联想记忆存储器实现与研究[D].安徽大学,2010,(10).
[44]王剑.一类新型联想记忆神经网络研究[D].华南理工大学,2003.