在自媒体时代,信息来源更多、扩散速度更快、扩散范围更广,这使得网络舆论的监控和管理更加困难。为提高网络舆情的管控能力,需深入了解网络中信息的扩散过程及重要特征。本文首先从微观角度分析信息的扩散机理,考虑网络的结构特点以及舆情监控的时效性,引入与用户是否在线相对应的节点状态,将传统的独立级联模型扩展为基于离散时间的双概率独立级联扩散模型。接着,本文从宏观角度对信息的扩散过程进行分析并定量表示,结合信息自身质量和用户网络特征两个客观要素,并考虑外部平台的影响,进而建立有关事件的动态扩散方程。与以往研究只利用模型模拟或纯粹用数值拟合相比,本文实验给出了社会网络中的信息扩散过程的一个更精准的刻画。实验结果展现了信息传播的多种形态,同时发现用户规模越大、用户关联越紧密以及信息质量越高时,信息爆发所需的时间越短。这些发现有利于预测信息扩散的趋势,同时为舆情管控的时效性和网络用户群体提供参考。
With more sources, higher speed and broader spreading, the public opinions in the network are more difficult to be monitored and controlled. In order to improve the ability of controlling the online public opinions, it is necessary to model the process of information diffusion quantitatively so that we can understand the characteristics of information dissemination. In this paper' the mechanism of information diffusion is first analyzed from the micro perspective, considering both the structural characteristics of networks and the time efficiency of controlling. By taking the state of node into consideration, which corresponds to a user's online or offline, the traditional independent cascade mode is developed into discrete-time oriented bi-probability cascade model. Then, the process of information diffusion is analyzed and quantified from the macro perspective. The dynamic diffusion equation with the quality of information, features of the network and the impact from external platforms is established. Compared with previous studies using only model simulations or numerical fitting, the process of information diffusion in social networks is descrbed more accurately. The experimental results show various forms of information dissemination with different characteristics. Besides, it also shows that less time would be taken for information to burst out with larger size of user network, closer relationship of users and higher quality of information. These findings are helpful to predict the trend of information diffusion and make it easier to recognize important propagators in time when managing the public opinions in social networks.
[1] Kempe D, Kleinberg J, Tardos E. Maximizing the spread of influence through a social network[C]//Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining,Washington D C,August,24-27,2003.
[2] Saito K, Kimura M, Ohara K, et al. Learning continuous-time information diffusion model for social behavioral data analysis[M]//Advances in Machine Learning. Springer Berlin Heidelberg, 2009:322-337.
[3] Saito K, Kimura M, Ohara K, et al. Selecting information diffusion models over social networks for behavioral analysis[M]//Zhou Zhihua,Washio T.Machine learning and knowledge discovery in databases. Berlin-Heidelberg:Springer,2010:180-195.
[4] 万圣贤, 郭嘉丰, 兰艳艳, 等. 基于传播模拟的消息流行度预测[J]. 中文信息学报, 2014,28(3):68-74.
[5] Chen Wei, Wang Yajun, & Yang Siyu. Efficient influence maximization in social networks[C]//proceedings of the 15th ACM SIGKDD International Conference on Knowledge discovery and Data Mining,Paris,France,June 28-July 1,2009.
[6] 朱湘, 贾焰, 聂原平. 基于微博的事件扩散分析[J]. 计算机研究与发展, 2015, 52(2):437-444.
[7] Saito K, Nakano R, Kimura M. Prediction of information diffusion probabilities for independent cascade model[M]//Lovrek I, Howlett R J,Jian L C.Knowledge-based intelligent information and engineering systems. Berlin-Heidelberg:Springer, 2008:67-75.
[8] Goyal A, Bonchi F, Lakshmanan L V S. Learning influence probabilities in social networks[C]//Proceedings of the third ACM international conference on Web search and data mining,New York,us,February 04-06,2010.
[9] 郭静, 曹亚男, 周川. 基于线性阈值模型的影响力扩散权重学习[J]. 电子与信息学报, 2014, 36(8):1804-1809.
[10] Srivastava J,Pathak N, Banerjee A. A generalized linear threshold model for multiple cascades[C]//Proceedings of the 13th International Conference on Data Ming,December13-17,2010.
[11] He Xinran, Songuojie G, Chen Wei, et al. Influence blocking maximization in social networks under the competitive linear threshold model[J]//Compting Science,2011.
[12] Litou I, Kalogeraki V, Katakis I. Real-time and cost-effective limitation of misinformation propagation[C]//Proceedings of the 17th TEEE Internation Confernce on Mobile Data Management,Porto,Portugal,June 13-16,2016.
[13] 田家堂,王轶彤,冯小军. 一种新型的社会网络影响最大化算法[J]. 计算机学报,2011,34(10):1956-1965.
[14] Yang J, Leskovec J. Patterns of temporal variation in online media[C]//Proceedings of the fourth ACM international conference on Web search and data mining,Hong kong,China,February 09-11,2011.
[15] 刘德海, 苏烨, 王维国. 振荡型群体性突发事件中信息特征的演化博弈分析[J].中国管理科学,2012,20(S1):172-178.
[16] Boyd D, Golder S, Lotan G. Tweet, tweet, retweet:Conversational aspects of retweeting on twitter[C]//the 43rd Hawaii International Conference on System Sciences (HICSS),Honolulu,HI,USA,January5-8,2010.
[17] 廖卫民, 柯伟. 网络舆论波研究——基于波浪力学及杭州两起舆论事件的理论思考[J]. 新闻记者, 2010,(4):12-16.
[18] 刘樑, 戴伟, 李仕明. 基于多Agent的非常规突发事件在线信息预警策略研究[J]. 中国管理科学, 2014,22(S1):180-187.
[19] 王秀利, 朱建明. 社会舆论方向影响下的微博商业言论传播模型[J]. 中国管理科学, 2012,20(S2):691-695.
[20] 曹学艳, 张仙, 刘樑,等. 基于应对等级的突发事件网络舆情热度分析[J]. 中国管理科学, 2014, 22(3):82-89.
[21] Matsubara Y, Sakurai Y, Prakash B A, et al. Rise and fall patterns of information diffusion:model and implications[C]//Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining. Beijing,August 12-16,2012.