Spatio-temporal Feature Based Emotional Contagion Analysis and Prediction Model for Online Social Networks
-
摘要: 社交网络用户情绪传播与用户的空间距离和时间跨度有关，并且受到多种交互机制的影响.从大规模社交网络数据中提取情绪传播的时空特征，研究用户行为对情绪传播的影响，对预测情绪传播趋势具有实际意义.利用线性回归获取的各行为子层的情绪传输率之间存在差异.提出一种基于多层社交网络的情绪传播模型，被称为ECM模型（Emotional contagion model）.该模型包括三个行为子层，每层的拓扑结构各不相同，由该行为的交互历史决定.在真实数据上对ECM模型进行仿真分析，可以获得社交网络中情绪传播的过程与规律：1）中性情绪用户所占比例随时间逐渐增大，接近57.1%，而正向情绪与负向情绪比例始终接近.2）情绪传输率越大，用户情绪更容易受到其他用户的影响而发生变化；初始情绪越中立的用户，在演化过程中情绪波动越小，而初始情绪极性越大的用户情绪波动越大.此外，通过实验对比该模型与其他情绪传播模型，表明ECM模型更加接近真实数据，对社交网络中情绪传播具有较好的预测效果，预测准确率相比其他模型可以提高1.8%~7.8%.Abstract: Users' emotion in social networks is related to spatial distance and time span, and affected by multiple interaction mechanisms. It has practical significance to extract the spatio-temporal features from large-scale social networks and study the influence of users' behaviors on emotional contagion in order to predict the trend of emotional contagion. The transmisibility values on different behavioral layers are calculated by linear regression and the results show the differences between these values. An emotional contagion model called ECM on multilayer social networks is proposed. It consists of three behavioral layers with different topologies depending on users' interaction history. By simulation on real dataset, it is discovered that, 1) the proportion of users with neutral emotion is gradually increased with time and reaches 57.1% while the proportion of positive emotion is comparable to that of negative emotion from beginning to end; 2) users' emotion is more likely to be influenced by other users when transmissibility becomes larger and users with initial polar emotion fluctuate more drastically than users with initial neutral emotion. In order to show the advantages of the proposed model, it is compared with other emotional contagion models. The results demonstrate that the proposed model approximates to the real data of emotional contagion on social networks, and also shows better predictive performance of emotional contagion. The prediction accuracy is increased by 1.8%~7.8%.
-
Key words:
- Emotion contagion /
- multilayer networks /
- behavior analysis /
- social networks
1) 本文责任编委 赵铁军 -
图 1 社交网络中情绪传播分析及模型构建示意图
Fig. 1 Analysis and modeling of emotion contagion in social networks
图 4 情绪转换数随用户初始情绪与节点度乘积的变化
Fig. 4 The relation between the number of individual emotional tendency changes, the degree and the initial emotion
图 5 三种模型与真实数据的对比(Twitter数据集)
Fig. 5 The comparison of the three models and the real data (Twitter dataset)
图 6 三种模型分类度量值的对比
Fig. 6 The comparison of classification measurements of the three models
图 7 三种模型中的F-1值随$\theta_1$的变化规律(Twitter数据集)
Fig. 7 F-1 changes with $\theta_1$ for the three models (Twitter dataset)
图 8 三种模型中的F-1值随用户数的变化规律(Twitter数据集)
Fig. 8 F-1 changes with the number of users for the three models (Twitter dataset)
表 1 数据集统计信息
Table 1 The statistical information of the datasets
Higgs数据集 数据堂数据集 新采集Twitter 新采集新浪微博 数据来源 Twitter 新浪微博 Twitter 新浪微博 用户(节点)数 456 626 63 641 33 070 6 344 好友关系数 14 855 842 1 391 718 185 393 54 093 转发次数 328 132 27 759 88 677 24 027 提及次数 150 818 未采集 41 245 10 428 回复次数 32 523 未采集 12 174 4 207 是否包含文本 否 是 是 是 
下载: 导出CSV
表 2 两个数据集不同子层的情绪传输率
Table 2 The transimisibilities on different layers in the two datasets
Twitter数据集 新浪微博数据集 转发 0.27 0.31 提及 0.95 1.07 回复 0.44 0.45
下载: 导出CSV
-
[1] Gabriel A S, Cheshin A, Moran C M, van Kleef G A. Enhancing emotional performance and customer service through human resources practices:a systems perspective. Human Resource Management Review, 2016, 26(1):14-24 http://www.sciencedirect.com/science/article/pii/S1053482215000510 [2] Xiong X B, Zhou G, Huang Y Z, Chen H Y, Xu K. Dynamic evolution of collective emotions in social networks:a case study of Sina weibo. Science China Information Sciences, 2013, 56(7):1-18 doi: 10.1007/s11432-013-4892-8 [3] Lo S C, Huang K P. The smiling mask in service encounters:the impact of surface and deep acting. International Journal of Management, Economics and Social Sciences, 2017, 6(1):40-55 http://www.econstor.eu/handle/10419/157922 [4] Wang Q Y, Lin Z, Jin Y H, Cheng S D, Yang T. ESIS:emotion-based spreader-ignorant-stifler model for information diffusion. Knowledge-Based Systems, 2015, 81:46-55 doi: 10.1016/j.knosys.2015.02.006 [5] Wang Q Y, Jin Y H, Yang T, Cheng S D. An emotion-based independent cascade model for sentiment spreading. Knowledge-Based Systems, 2017, 116:86-93 doi: 10.1016/j.knosys.2016.10.029 [6] Dignath D, Janczyk M, Eder A B. Phasic valence and arousal do not influence post-conflict adjustments in the Simon task. Acta Psychologica, 2017, 174:31-39 doi: 10.1016/j.actpsy.2017.01.004 [7] Lloyd-Jones D M, Larson M G, Leip E P, Beiser A, D'Agostino R B, Kannel W B, et al. Lifetime risk for developing congestive heart failure:the Framingham Heart Study. Circulation, 2002, 106(24):3068-3072 doi: 10.1161/01.CIR.0000039105.49749.6F [8] Coviello L, Sohn Y, Kramer A D I, Marlow C, Franceschetti M, Christakis N A, et al. Detecting emotional contagion in massive social networks. PLoS One, 2014, 9(3): Article No. e90315 [9] Del Vicario M, Vivaldo G, Bessi A, Zollo F, Scala A, Caldarelli G, et al. Echo chambers: emotional contagion and group polarization on Facebook. Scientific Reports, 2016, 6: Article No. 37825 [10] Guille A, Hacid H, Favre C, Zighed D A. Information diffusion in online social networks:a survey. ACM SIGMOD Record, 2013, 42(2):17-28 http://d.old.wanfangdata.com.cn/OAPaper/oai_arXiv.org_1310.0505 [11] Bozorgi A, Samet S, Kwisthout J, Wareham T. Community-based influence maximization in social networks under a competitive linear threshold model. Knowledge-Based Systems, 2017, 134:149-158 doi: 10.1016/j.knosys.2017.07.029 [12] Liu L J. A delayed SIR model with general nonlinear incidence rate. Advances in Difference Equations, 2015, 2015:329 doi: 10.1186/s13662-015-0619-z [13] Zhang X H, Jiang D Q, Alsaedi A, Hayat T. Stationary distribution of stochastic SIS epidemic model with vaccination under regime switching. Applied Mathematics Letters, 2016, 59:87-93 doi: 10.1016/j.aml.2016.03.010 [14] Xiong F, Liu Y, Zhang Z J, Zhu J, Zhang Y. An information diffusion model based on retweeting mechanism for online social media. Physics Letters A, 2012, 376(30):2103-2108 http://www.sciencedirect.com/science/article/pii/S037596011200566X [15] Boccaletti S, Bianconi G, Criado R, del Genio C I, Gómez-Gardeñes J, Romance M, et al. The structure and dynamics of multilayer networks. Physics Reports, 2014, 544(1):1-122 doi: 10.1016/j.physrep.2014.07.001 [16] KivelaM, Arenas A, Barthelemy M, Gleeson J P, Moreno Y, Porter M A. Multilayer networks. Journal of Complex Networks, 2013, 2(3):261-268 http://d.old.wanfangdata.com.cn/Periodical/jsjyjyfz200606008 [17] Yaǧan O, Qian D J, Zhang J S, Cochran D. Information diffusion in overlaying social-physical networks. In: Proceedings of the 201246th Annual Conference on Information Sciences and Systems (CISS). Princeton, NJ, USA: IEEE, 2012. 1038-1048 [18] Zhuang Y, Yagan O. Information propagation in clustered multilayer networks. IEEE Transactions on Network Science and Engineering, 2016, 3(4):211-224 doi: 10.1109/TNSE.2016.2600059 [19] Kim M, Newth D, Christen P. Modeling dynamics of diffusion across heterogeneous social networks:news diffusion in social media. Entropy, 2013, 15(15):4215-4242 http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2255707 [20] Xiong X, Li Y Y, Qiao S J, Han N, Wu Y, Peng J, et al. An emotional contagion model for heterogeneous social media with multiple behaviors. Physica A:Statistical Mechanics and its Applications, 2018, 490:185-202 doi: 10.1016/j.physa.2017.08.025 [21] Ferrara E, Yang Z Y. Measuring emotional contagion in social media. PLoS One, 2015, 10(11): Article No. e0142390 [22] Kramer A D I. The spread of emotion via Facebook. in: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. Austin, Texas, USA: ACM, 2012. 767-770 [23] De Domenico M, Lima A, Mougel P, Musolesi M. The anatomy of a scientific rumor. Scientific Reports, 2013, 3: Article No. 2980 [24] Thelwall M, Buckley K, Paltoglou G. Sentiment strength detection for the social web. Journal of the Association for Information Science and Technology, 2012, 63(1):163-173 doi: 10.1002/asi.21662/full [25] 张云.基于开源软件的中文学术文献计量软件的开发实践.现代图书情报技术, 2010, 26(4):87-91 http://www.cnki.com.cn/Article/CJFDTOTAL-XDTQ201004023.htmZhang Yun. The practice on the development of software on the Chinese academic bibliometrics based on the open source software. New Technology of Library and Information Service, 2010, 26(4):87-91 http://www.cnki.com.cn/Article/CJFDTOTAL-XDTQ201004023.htm [26] The Boson Data. Sentiment analysis[Online], available: http://docs.bosonnlp.com/sentiment.html, 2017. [27] Powers D M W. Evaluation:from precision, recall and F-measure to ROC, informedness, markedness & correlation. Journal of Machine Learning Technologies, 2011, 2(1):37-63 -
图(8) / 表(2)
计量
- 文章访问数: 2646
- HTML全文浏览量: 594
- PDF下载量: 670
- 被引次数: 0
