搜索

x

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

具有层级结构集体影响力的多数投票模型

陈奕多 韵雨婷 关剑月 吴枝喜

引用本文:
Citation:

具有层级结构集体影响力的多数投票模型

陈奕多, 韵雨婷, 关剑月, 吴枝喜

Majority-vote model with collective influence of hierarchical structures

Chen Yi-Duo, Yun Yu-Ting, Guan Jian-Yue, Wu Zhi-Xi
PDF
HTML
导出引用
  • 多数投票模型是观点动力学研究中的常用模型, 本文在多数投票模型的基础上引入了具有层级结构的集体影响力, 以节点周边层级结构上的节点的度衡量中心节点的观点权重, 即为集体影响力参数. 通过蒙特卡罗模拟, 研究了具有集体影响力的多数投票模型在ER (Erdos and Rényi)随机网络与无标度网络上观点的演化, 发现系统观点均出现了有序-无序相变, 且相比原始多数投票模型更容易趋于无序, 即相变临界点更小. 原因是考虑具有层级结构的集体影响力时, 系统的集体影响力参数值整体减小, 且分布数目随着参数值的增大而减少, 呈“长尾”趋势, 占少数的高影响力个体使周围节点的观点产生跟随现象, 随着噪声参数的增大, 当少数的高影响力个体趋于无序时, 整个系统也会趋于无序, 即系统更容易达到无序状态. 最后通过有限尺寸标度法, 发现无论在ER随机网络或在无标度网络中, 具有集体影响力的多数投票模型的相变均为Ising模型普适类.
    Majority-vote model is a commonly used model in the study of opinion dynamics. In the original majority-vote model, the influence of node is determined by their neighbors. But there are nodes with low degree surrounded by nodes with high degree so they also have a great influence on the evolution of opinions. Therefore, the influence of a node should not only be measured by neighbors but also be connected to itself directly. Thus, this paper adds collective influence with hierarchical structures into the majority-vote model and measures opinion weight of center node by degree of their neighbors on hierarchical structures surround it with the set distance. The collective influence parameters used in this paper are related to the value of collective influence mentioned above and normalized by the maximum value of all nodes in system. The opinions’ evolution of majority-vote model with collective influence is studied in ER random networks and scale-free networks with different degree distribution exponents by Monte Carlo simulations. It is found that all systems have order-to-disorder phase transitions with the increase of noise parameter. When the depth of hierarchical structure is not zero, the system with collective influence is much easier to turn to disordered states so their critical noise parameters of phase transition are smaller than those of 0-depth systems and original majority-vote model. The reason is that high degree nodes in original majority-vote model have high influence because they are connected to more neighbors and nodes’ influence is also directly determined by degree in 0-depth collective influence model. Furthermore, nodes’ collective influence parameters in the system will all decrease when hierarchical structure of nonzero depth is considered, only a small number of individuals have high influence parameters in the system and they will make the opinions of surrounding individuals follow theirs, so if the opinions of a few highly influential individuals are out of order, then the system will reach a state of disorder. Because of the above factors, the collective influence model of nonzero depth is much easier to disorder with the increase of noise parameter. Besides, the system proves to be easier to reach a disordered state with the increase of degree distribution exponents in scale-free networks because all nodes’ degree will be lower so that the system will be dominated by less nodes with high degree. This conclusion verifies that scale-free networks are more similar to ER random networks with the increase of degree distribution exponents. Finally, through the finite-size scaling method, it is found that the phase transition of the majority-vote model with collective influence of hierarchical structures belongs in the Ising model universal class, whether in ER random networks or in scale-free networks.
      通信作者: 关剑月, guanjy@lzu.edu.cn ; 吴枝喜, wuzhx@lzu.edu.cn
    • 基金项目: 国家自然科学基金(批准号: 11975111, 12047501, 12247101)资助的课题.
      Corresponding author: Guan Jian-Yue, guanjy@lzu.edu.cn ; Wu Zhi-Xi, wuzhx@lzu.edu.cn
    • Funds: Project supported by the National Natural Science Foundation of China (Grant Nos. 11975111, 12047501, 12247101).
    [1]

    Calvelli M, Crokidakis N, Penna T J P 2019 Physica A 513 518Google Scholar

    [2]

    Pires M A, Crokidakis N 2022 Phil. Trans. R. Soc. A. 380 20210164Google Scholar

    [3]

    Khalil N, Galla T 2021 Phys. Rev. E 103 012311Google Scholar

    [4]

    Liu J Z, Huang S D, Aden N M, Johnson N F, Song C M 2023 Phys. Rev. Lett. 130 037401Google Scholar

    [5]

    Costa L S A, de Souza A J F 2005 Phys. Rev. E 71 056124Google Scholar

    [6]

    Lima F W S, Fulco U L, Costa Filho R N 2005 Phys. Rev. E 71 036105Google Scholar

    [7]

    de Oliveira M J 1992 J. Stat. Phys. 66 273Google Scholar

    [8]

    Mendes J F F, Santos M A 1998 Phys. Rev. E 57 108Google Scholar

    [9]

    de Silva Hilho A G, Moreira F G B 2002 J. Stat. Phys. 106 391Google Scholar

    [10]

    Campos P R A, de Oliveira V M, Brady Moreira F G 2003 Phys. Rev. E 67 026104Google Scholar

    [11]

    Pereira L F C, Moreira F G B 2005 Phys. Rev. E 71 016123Google Scholar

    [12]

    Kwak W, Yang J S, Sohn J I, et al. 2007 Phys. Rev. E 75 061110Google Scholar

    [13]

    de Oliveira M J, Mendes J F F, Santos M A 1993 J. Phys. A 26 2317Google Scholar

    [14]

    Marques M C 1993 J. Phys. A 26 1559Google Scholar

    [15]

    Pastor-Satorras R, Vespignani A 2001 Phys. Rev. Lett. 86 3200Google Scholar

    [16]

    Lloyd A L, May R M 2001 Science 292 1316Google Scholar

    [17]

    Wang S J, Wu Z X, Dong H R, Chen G R 2010 Int. J. Mod. Phys. C 21 67Google Scholar

    [18]

    Morone F, Makse H A 2015 Nature 524 65Google Scholar

    [19]

    Binder K 1981 Physica B 43 119Google Scholar

    [20]

    普利施克 M, 贝格森 B著(汤雷翰, 童培庆译)2020 平衡态统计物理学(北京: 北京大学出版社)第174—177页

    Plischke M, Bergersen B (translated by Tang L H, Tong P Q) 2020 Equilibrium Statistical Physics (Beijing: Peking University Press) pp174–177

    [21]

    Molloy M, Reed B 1995 Random Struct. Algorithms 6 161Google Scholar

    [22]

    Molloy M, Reed B 1998 Comb. Probab. Comput. 7 295Google Scholar

    [23]

    Lima F W S 2006 Int. J. Mod. Phys. C 17 1257Google Scholar

    [24]

    Kim M, Yook S H 2021 Phys. Rev. E 103 022302Google Scholar

  • 图 1  (a), (b)分别表示在 ER 随机网络中磁化强度$ M(q, N) $、四阶宾德累积矩$ U(q, N) $随噪声参数$ q $的变化曲线, 网络平均度$ \left\langle{k}\right\rangle=10 $, 节点数$ N=10000 $

    Fig. 1.  (a), (b) The variation curves of magnetization $ M(q, N) $ and Binder’s fourth-order cumulant $ U(q, N) $ with noise parameter $ q $ in ER random network, respectively. The average degree of networks is $ \left\langle{k}\right\rangle=10 $, and the number of nodes is $ N=10000 $.

    图 2  (a), (b)在 ER 网络中, 当$ l=0 $和$ l=1 $时四阶宾德累积矩$ U(q, N) $随噪声参数$ q $的变化曲线. 网络平均度$ \left\langle{k}\right\rangle=10 $

    Fig. 2.  (a), (b) $ l=0 $ and $ l=1 $ of the Binder’s fourth-order cumulant $ U(q, N) $ with noise parameter $ q $ in the ER network, respectively. The average degree of networks is $ \left\langle{k}\right\rangle=10 $.

    图 3  (a), (b)分别为ER随机网络中$ {\omega }_{i} $分布情况与不同度$ {k}_{i} $的$ {\omega }_{i} $平均值大小统计情况

    Fig. 3.  (a) The distribution of $ {\omega }_{i} $ and (b) the statistics of the average value of $ {\omega }_{i} $ of different degrees $ {k}_{i} $ in ER random network.

    图 4  (a), (b) 分别为ER网络在$ l=0 $时磁化强度$ M(q, N) $与磁化率$ \chi (q, N) $的有限尺寸标度图; (c), (d)分别为ER网络在$ l=1 $时磁化强度$ M(q, N) $与磁化率$ \chi (q, N) $的有限尺寸标度图

    Fig. 4.  (a), (b) The finite size scaling graphs of the magnetization $ M(q, N) $ and the susceptibility $ \chi (q, N) $ of the ER network at $ l=0 $, respectively; (c), (d) the finite size scaling graphs of the magnetization $ M(q, N) $ and the susceptibility $ \chi (q, N) $ of the ER network at $ l= 1 $, respectively.

    图 5  (a), (b) $ \lambda =2.5 $; (c), (d) $ \lambda =3 $; (e), (f) $ \lambda =4 $无标度网络中磁化强度$ M(q, N) $、四阶宾德累积矩$ U(q, N) $随噪声参数$ q $的变化曲线. 网络平均度$ \left\langle{k}\right\rangle=10 $, 节点数$ N=10000 $

    Fig. 5.  The variation curves of magnetization $ M(q, N) $ and Binder’s fourth-order cumulant $ U(q, N) $ with noise parameter $ q $ in scale-free networks with (a), (b) $ \lambda =2.5 $; (c), (d) $ \lambda =3 $; (e), (f) $ \lambda =4 $, respectively. The average degree of networks is $ \left\langle{k}\right\rangle=10 $, and the number of nodes is $ N=10000 $.

    图 6  (a), (b) $ \lambda =2.5 $; (c), (d) $ \lambda =3 $; (e), (f) $ \lambda =4 $无标度网络中$ l=0 $和$ l=1 $时四阶宾德累积矩$ U(q, N) $随噪声参数$ q $的变化曲线. 网络平均度$ \left\langle{k}\right\rangle=10 $

    Fig. 6.  The variation curves of the Binder’s fourth-order cumulant $ U(q, N) $ with the noise parameter q when $ l=0 $ and $ l=1 $ in a scale-free network with (a), (b) $ \lambda =2.5 $; (c), (d) $ \lambda =3 $; (e), (f) $ \lambda =4 $, respectively. The average degree of networks is $ \left\langle{k}\right\rangle=10 $.

    图 7  $ \lambda =2.5 $时无标度网络磁化强度$ M(q, N) $和磁化率$ \chi (q, N) $的有限尺寸标度图 (a), (b) $ l=0 $; (c), (d) $ l=1 $

    Fig. 7.  The finite-size scaling graphs of magnetization $ M(q, N) $ and susceptibility $ \chi (q, N) $ of scale-free networks with $ \lambda =2.5 $: (a), (b) $ l=0 $; (c), (d) $ l=1 $.

    图 8  $ l=0 $和$ l=1 $的具有集体影响力的多数投票模型中, 无标度网络不同度分布指数$ \lambda $对应的相变临界点$ {q}_{{\mathrm{c}}} $

    Fig. 8.  In the majority-vote model with collective influence of $ l=0 $ and $ l=1 $, the phase transition critical point $ {q}_{{\mathrm{c}}} $ with different degree distribution index $ \lambda $ of the scale-free network.

    图 9  $ N=10000 $, $ \lambda =2.5 $的无标度网络中 (a) $ {\omega }_{i} $的分布情况; (b)不同度$ {k}_{i} $的$ {\omega }_{i} $平均值统计情况

    Fig. 9.  In scale-free network when $ N=10000 $ and $ \lambda =2.5 $: (a) Distribution of $ {\omega }_{i} $ in scale-free networks; (b) the statistics of the average value of $ {\omega }_{i} $ of different degrees $ {k}_{i} $ .

    图 10  $ {\omega }_{i} $的分布情况图, $ \lambda =2.5 $的无标度网络在$ l=0 $和$ l=1 $时的分布情况与$ \lambda =4 $的无标度网络在$ l=0 $和$ l=1 $时的分布情况

    Fig. 10.  Distribution of $ {\omega }_{i} $, $ l=0 $, $ l=1 $ of scale-free networks with $ \lambda =2.5 $ and $ \lambda =4 $.

    图 11  相变临界点处的磁化强度值. 横坐标为系统尺寸大小

    Fig. 11.  The magnetization $ M(q, N) $ value at the critical point of phase transition. The abscissa is the size of the system.

    表 1  $ {\omega }_{i} $均值和方差

    Table 1.  Mean and variance of $ {\omega }_{i} $.

    $ l=0 $$ l=1 $$ l=2 $$ l=3 $
    $ \left\langle{{\omega }_{i}}\right\rangle $$ 0.41\left(6\right) $$ 0.18\left(5\right) $$ 0.18\left(3\right) $$ 0.19\left(3\right) $
    $ \sigma \left({\omega }_{i}\right) $$ 0.017\left(4\right) $$ 0.014\left(9\right) $$ 0.014\left(5\right) $$ 0.014\left(9\right) $
    下载: 导出CSV

    表 2  不同网络对应的$ {q}_{{\mathrm{c}}0} $, $ {q}_{{\mathrm{c}}1} $以及$ |{q}_{{\mathrm{c}}0}-{q}_{{\mathrm{c}}1}| $

    Table 2.  $ {q}_{{\mathrm{c}}0} $, $ {q}_{{\mathrm{c}}1} $ and $ |{q}_{{\mathrm{c}}0}-{q}_{{\mathrm{c}}1}| $ of different networks.

    ER网络$ \lambda =2.5 $$ \lambda =2.7 $$ \lambda =3 $$ \lambda =3.5 $$ \lambda =4 $
    $ {q}_{{\mathrm{c}}0} $0.3010.30350.30150.30.2970.295
    $ {q}_{{\mathrm{c}}1} $0.2830.2920.28950.28750.28350.2805
    $ |{q}_{{\mathrm{c}}0}-{q}_{{\mathrm{c}}1}| $0.0180.01150.0120.01250.01350.0145
    下载: 导出CSV

    表 3  $ {\omega }_{i} $均值和方差

    Table 3.  Mean and variance of $ {\omega }_{i} $.

    $ l=0 $$ l=1 $$ l=2 $$ l=3 $
    $ \left\langle{{\omega }_{i}}\right\rangle $0.66(2)0.37(6)0.36(9)0.38(1)
    $ \sigma \left({\omega }_{i}\right) $0.22(4)0.035(2)0.033(8)0.033(8)
    下载: 导出CSV

    表 4  $ {\omega }_{i} $均值和方差

    Table 4.  Mean and variance of $ {\omega }_{i} $.

    $ l=0 $$ l=1 $$ l=2 $$ l=3 $
    $ \left\langle{{\omega }_{i}}\right\rangle $$ \lambda =2.5 $0.66(2)0.37(6)0.36(9)0.38(1)
    $ \lambda =4 $0.61(7)0.33(6)0.33(7)0.34(6)
    $ \sigma \left({\omega }_{i}\right) $$ \lambda =2.5 $0.022(5)0.035(2)0.033(8)0.033(8)
    $ \lambda =4 $0.017(4)0.027(4)0.027(5)0.027(4)
    下载: 导出CSV

    表 5  临界指数实验结果与引用数据对照

    Table 5.  Results of critical exponents and reference data for comparison.

    $ l=0 $ ER网络 $ \lambda =2.5 $ $ \lambda =3 $ $ \lambda =4 $
    $ 1/\bar{\nu } $0.49(6)0.45(5)0.46(1)0.48(5)
    $ \beta /\bar{\nu } $0.23(5)0.23(1)0.23(5)0.23(5)
    $ \gamma /\bar{\nu } $0.49(5)0.49(2)0.49(2)0.49(5)
    $ l=1 $ER网络$ \lambda =2.5 $$ \lambda =3 $$ \lambda =4 $
    $ 1/\bar{\nu } $0.47(5)0.44(6)0.46(5)0.47(5)
    $ \beta /\bar{\nu } $0.23(6)0.22(1)0.23(1)0.23(5)
    $ \gamma /\bar{\nu } $0.50(5)0.51(5)0.51(5)0.51(2)
    原始多数投票模型ER网络$ \lambda < 3 ~ (\lambda =2.7) $$ 3 < \lambda < 5 ~ (\lambda =3.7) $$ \lambda > 5~ (\lambda =5.2) $
    $ 1/\bar{\nu } $$ 0.5 $$ 0.31 $$ 0.48 $$ 0.47 $
    $ \beta /\bar{\nu } $$ 0.25 $$ 0.25 $$ 0.25 $$ 0.21 $
    $ \gamma /\bar{\nu } $$ 0.5 $$ 0.51 $$ 0.49 $$ 0.57 $
    下载: 导出CSV
  • [1]

    Calvelli M, Crokidakis N, Penna T J P 2019 Physica A 513 518Google Scholar

    [2]

    Pires M A, Crokidakis N 2022 Phil. Trans. R. Soc. A. 380 20210164Google Scholar

    [3]

    Khalil N, Galla T 2021 Phys. Rev. E 103 012311Google Scholar

    [4]

    Liu J Z, Huang S D, Aden N M, Johnson N F, Song C M 2023 Phys. Rev. Lett. 130 037401Google Scholar

    [5]

    Costa L S A, de Souza A J F 2005 Phys. Rev. E 71 056124Google Scholar

    [6]

    Lima F W S, Fulco U L, Costa Filho R N 2005 Phys. Rev. E 71 036105Google Scholar

    [7]

    de Oliveira M J 1992 J. Stat. Phys. 66 273Google Scholar

    [8]

    Mendes J F F, Santos M A 1998 Phys. Rev. E 57 108Google Scholar

    [9]

    de Silva Hilho A G, Moreira F G B 2002 J. Stat. Phys. 106 391Google Scholar

    [10]

    Campos P R A, de Oliveira V M, Brady Moreira F G 2003 Phys. Rev. E 67 026104Google Scholar

    [11]

    Pereira L F C, Moreira F G B 2005 Phys. Rev. E 71 016123Google Scholar

    [12]

    Kwak W, Yang J S, Sohn J I, et al. 2007 Phys. Rev. E 75 061110Google Scholar

    [13]

    de Oliveira M J, Mendes J F F, Santos M A 1993 J. Phys. A 26 2317Google Scholar

    [14]

    Marques M C 1993 J. Phys. A 26 1559Google Scholar

    [15]

    Pastor-Satorras R, Vespignani A 2001 Phys. Rev. Lett. 86 3200Google Scholar

    [16]

    Lloyd A L, May R M 2001 Science 292 1316Google Scholar

    [17]

    Wang S J, Wu Z X, Dong H R, Chen G R 2010 Int. J. Mod. Phys. C 21 67Google Scholar

    [18]

    Morone F, Makse H A 2015 Nature 524 65Google Scholar

    [19]

    Binder K 1981 Physica B 43 119Google Scholar

    [20]

    普利施克 M, 贝格森 B著(汤雷翰, 童培庆译)2020 平衡态统计物理学(北京: 北京大学出版社)第174—177页

    Plischke M, Bergersen B (translated by Tang L H, Tong P Q) 2020 Equilibrium Statistical Physics (Beijing: Peking University Press) pp174–177

    [21]

    Molloy M, Reed B 1995 Random Struct. Algorithms 6 161Google Scholar

    [22]

    Molloy M, Reed B 1998 Comb. Probab. Comput. 7 295Google Scholar

    [23]

    Lima F W S 2006 Int. J. Mod. Phys. C 17 1257Google Scholar

    [24]

    Kim M, Yook S H 2021 Phys. Rev. E 103 022302Google Scholar

  • [1] 李江, 刘影, 王伟, 周涛. 识别高阶网络传播中最有影响力的节点. 物理学报, 2024, 73(4): 048901. doi: 10.7498/aps.73.20231416
    [2] 孔江涛, 黄健, 龚建兴, 李尔玉. 基于复杂网络动力学模型的无向加权网络节点重要性评估. 物理学报, 2018, 67(9): 098901. doi: 10.7498/aps.67.20172295
    [3] 苏臻, 高超, 李向华. 节点中心性对复杂网络传播模式的影响分析. 物理学报, 2017, 66(12): 120201. doi: 10.7498/aps.66.120201
    [4] 阮逸润, 老松杨, 王竣德, 白亮, 侯绿林. 一种改进的基于信息传播率的复杂网络影响力评估算法. 物理学报, 2017, 66(20): 208901. doi: 10.7498/aps.66.208901
    [5] 韩忠明, 陈炎, 李梦琪, 刘雯, 杨伟杰. 一种有效的基于三角结构的复杂网络节点影响力度量模型. 物理学报, 2016, 65(16): 168901. doi: 10.7498/aps.65.168901
    [6] 苏晓萍, 宋玉蓉. 利用邻域“结构洞”寻找社会网络中最具影响力节点. 物理学报, 2015, 64(2): 020101. doi: 10.7498/aps.64.020101
    [7] 闵磊, 刘智, 唐向阳, 陈矛, 刘三(女牙). 基于扩展度的复杂网络传播影响力评估算法. 物理学报, 2015, 64(8): 088901. doi: 10.7498/aps.64.088901
    [8] 胡庆成, 张勇, 许信辉, 邢春晓, 陈池, 陈信欢. 一种新的复杂网络影响力最大化发现方法. 物理学报, 2015, 64(19): 190101. doi: 10.7498/aps.64.190101
    [9] 袁铭. 带有层级结构的复杂网络级联失效模型. 物理学报, 2014, 63(22): 220501. doi: 10.7498/aps.63.220501
    [10] 王亚奇, 王静, 杨海滨. 基于复杂网络理论的微博用户关系网络演化模型研究. 物理学报, 2014, 63(20): 208902. doi: 10.7498/aps.63.208902
    [11] 苑卫国, 刘云, 程军军, 熊菲. 微博双向关注网络节点中心性及传播 影响力的分析. 物理学报, 2013, 62(3): 038901. doi: 10.7498/aps.62.038901
    [12] 李炎, 唐刚, 宋丽建, 寻之朋, 夏辉, 郝大鹏. Erds Rnyi随机网络上爆炸渗流模型相变性质的数值模拟研究. 物理学报, 2013, 62(4): 046401. doi: 10.7498/aps.62.046401
    [13] 李泽荃, 张瑞新, 杨曌, 赵红泽, 于健浩. 复杂网络中心性对灾害蔓延的影响. 物理学报, 2012, 61(23): 238902. doi: 10.7498/aps.61.238902
    [14] 吕业刚, 梁晓琳, 龚跃球, 郑学军, 刘志壮. 外加电场对铁电薄膜相变的影响. 物理学报, 2010, 59(11): 8167-8171. doi: 10.7498/aps.59.8167
    [15] 邢长明, 刘方爱. 基于Sierpinski分形垫的确定性复杂网络演化模型研究. 物理学报, 2010, 59(3): 1608-1614. doi: 10.7498/aps.59.1608
    [16] 王建伟, 荣莉莉. 基于负荷局域择优重新分配原则的复杂网络上的相继故障. 物理学报, 2009, 58(6): 3714-3721. doi: 10.7498/aps.58.3714
    [17] 欧阳敏, 费 奇, 余明晖. 基于复杂网络的灾害蔓延模型评价及改进. 物理学报, 2008, 57(11): 6763-6770. doi: 10.7498/aps.57.6763
    [18] 马卫东, 王 磊, 李幼平, 水鸿寿, 周明天. 用户需求行为对互联网动力学整体特性的影响. 物理学报, 2008, 57(3): 1381-1388. doi: 10.7498/aps.57.1381
    [19] 李 季, 汪秉宏, 蒋品群, 周 涛, 王文旭. 节点数加速增长的复杂网络生长模型. 物理学报, 2006, 55(8): 4051-4057. doi: 10.7498/aps.55.4051
    [20] 袁坚, 任勇, 刘锋, 山秀明. 复杂计算机网络中的相变和整体关联行为. 物理学报, 2001, 50(7): 1221-1225. doi: 10.7498/aps.50.1221
计量
  • 文章访问数:  2435
  • PDF下载量:  113
  • 被引次数: 0
出版历程
  • 收稿日期:  2023-07-19
  • 修回日期:  2023-10-05
  • 上网日期:  2023-10-12
  • 刊出日期:  2024-01-20

/

返回文章
返回