基于情感的社会网传播模型及影响最大化算法研究
|
摘要
针对社会网传播领域的影响最大化问题的研究,将节点本身具备的情感对事件传播的影响力进行了忽略,提出了基于情感的社会网传播模型(Emotion Independent Cascade model,E-IC),关于E-IC模型重点强调了情感影响的最大化基本问题(Influence Maximization Problem based on Emotion,IMPE),在传播整个进程中,融合了用户位置的计算值、后置情感的计算值以及交互概率值。论证并确认基于情感的社会网传播模型问题就是NP-hard问题,并给出近似算法EMS-Greedy。在训练集上调整模型参数,使得传播过程更符合传播规律,通过大规模真实数据集上的实验验证了E-IC模型的有效性。与其他模型相比,E-IC模型在传播范围上扩大了7%左右。
In view of the social network influence maximization problem in the field of communication research, the node itself has emotion to ignore the influence of the event propagation, this paper puts forward the social network communication model based on emotion, Emotion Independent Cascade model(E-IC). About E-IC model, this paper emphasizes the emotional impact of basic problems(Influence Maximization Problem based on Emotion, IMPE). In the entire spread process, this paper combines the calculated value of user's location, the calculated value of rear emotion and interactive probability value. This paper demonstrates and confirms that the social network communication model based on emotional problem is NP-hard problem, and gives EMS-Greedy approximate algorithm. The model parameters are adjusted on the training set to make the propagation process more consistent with the propagation law. The effectiveness of the E-IC model is verified by experiments on large-scale real data sets. Compared with other models, E-IC model expands its range by about 7%.
In view of the social network influence maximization problem in the field of communication research, the node itself has emotion to ignore the influence of the event propagation, this paper puts forward the social network communication model based on emotion, Emotion Independent Cascade model(E-IC). About E-IC model, this paper emphasizes the emotional impact of basic problems(Influence Maximization Problem based on Emotion, IMPE). In the entire spread process, this paper combines the calculated value of user's location, the calculated value of rear emotion and interactive probability value. This paper demonstrates and confirms that the social network communication model based on emotional problem is NP-hard problem, and gives EMS-Greedy approximate algorithm. The model parameters are adjusted on the training set to make the propagation process more consistent with the propagation law. The effectiveness of the E-IC model is verified by experiments on large-scale real data sets. Compared with other models, E-IC model expands its range by about 7%.
引文
[1] Kempe D,Kleinberg J,Tardos E.Maximizing fhe spread of influence through a social network[C]//Proceedings of the 9th ACM SIGKDD Intematiotial Conference on Knowledge Discovery and Data Mining,Washington,USA,Aug 24-27,2003:137-146.
[2] Tang Jie,Wu Sen,Sun Jimeng.Confluence:Conformity influence in large social networks[C]//Proceedings of the19th ACM SIGKDD International Conference on Knowledge Discovery and Data Minging,Chicago,USA,Aug11-14,2013:347-355.
[3] Li Guoliang,Chen Shuo,Feng Jianhua,et al.Efficient“location-aware influence maximization”[C]//Proceedings of the 2014 ACM SIGMOD International Conference on Management of Data,Snowbird,USA,Jun 22-27,2014:87-98.
[4]刘勇,曲思桐,王楠,等.不同营销模式中基于时间的影响传播方法研究[J].计算机科学与探索,2016,10(3):338-349.
[5] Domingos P,Richardson M.Mining the network value of customers[C]//Proceedings of the 7th ACM SIGKDD Conference on Knowledge Discovery and Data Mining,San Francisco,USA,Aug 26-29,2001:57-66.
[6] Kempe D,Kleinbetg J,Tardos E.Influential nodes in a diffusion model for social networks[C]//Proceedings of the 32nd International Colloquium on Automata,Languages and Programming,Lisbon,Portugal,Jul 11-15,2005:1127-1138.
[7] Chen W,Wang C,Wang Y.Scalable influence maximization for prevalent viral marketing in large-scale social networks[C]//Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining,2010:1029-1038.
[8] Galhotra S.Holistic influence maximization-combining scalability and efficiency with opinion-aware models[C]//Proceedings of the 2016 ACM SIGMOD International Conference on Management of Data,San Francisco,USA,2016:743-758.
[9]刘亚州,王静,潘晓中,等.考虑节点亲密度的社交网络谣言传播研究[J].计算机工程与应用,2018,54(14):100-106.
[10] Li Hang.Statistical learning methods[M]//The EM algorithm and its promotion.Beijing:Tsinghua Press,2012:155-168.
[11] Zhang H Y,Dinh T N,Thai M T.Maximizing the spread of positive influence in online social networks[C]//Proceedings of International Conference on Distributed Computing Systems,2013:317-326.
[12] Gao H,Mahmud J,Chen J,et al.Modeling user attitude toward controversial topics in online social media[C]//Proceedings of the 8th International Conference on Weblogs and Social Media,2014.
[13] Aslay C,Barbieri N,Bonchi F,et al.Online topic-aware influence maximization queries[C]//Proceedings of the17th International Conference on Extending Database Technology,Athens,Greece,Mar 24-28,2014:295-306.
[2] Tang Jie,Wu Sen,Sun Jimeng.Confluence:Conformity influence in large social networks[C]//Proceedings of the19th ACM SIGKDD International Conference on Knowledge Discovery and Data Minging,Chicago,USA,Aug11-14,2013:347-355.
[3] Li Guoliang,Chen Shuo,Feng Jianhua,et al.Efficient“location-aware influence maximization”[C]//Proceedings of the 2014 ACM SIGMOD International Conference on Management of Data,Snowbird,USA,Jun 22-27,2014:87-98.
[4]刘勇,曲思桐,王楠,等.不同营销模式中基于时间的影响传播方法研究[J].计算机科学与探索,2016,10(3):338-349.
[5] Domingos P,Richardson M.Mining the network value of customers[C]//Proceedings of the 7th ACM SIGKDD Conference on Knowledge Discovery and Data Mining,San Francisco,USA,Aug 26-29,2001:57-66.
[6] Kempe D,Kleinbetg J,Tardos E.Influential nodes in a diffusion model for social networks[C]//Proceedings of the 32nd International Colloquium on Automata,Languages and Programming,Lisbon,Portugal,Jul 11-15,2005:1127-1138.
[7] Chen W,Wang C,Wang Y.Scalable influence maximization for prevalent viral marketing in large-scale social networks[C]//Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining,2010:1029-1038.
[8] Galhotra S.Holistic influence maximization-combining scalability and efficiency with opinion-aware models[C]//Proceedings of the 2016 ACM SIGMOD International Conference on Management of Data,San Francisco,USA,2016:743-758.
[9]刘亚州,王静,潘晓中,等.考虑节点亲密度的社交网络谣言传播研究[J].计算机工程与应用,2018,54(14):100-106.
[10] Li Hang.Statistical learning methods[M]//The EM algorithm and its promotion.Beijing:Tsinghua Press,2012:155-168.
[11] Zhang H Y,Dinh T N,Thai M T.Maximizing the spread of positive influence in online social networks[C]//Proceedings of International Conference on Distributed Computing Systems,2013:317-326.
[12] Gao H,Mahmud J,Chen J,et al.Modeling user attitude toward controversial topics in online social media[C]//Proceedings of the 8th International Conference on Weblogs and Social Media,2014.
[13] Aslay C,Barbieri N,Bonchi F,et al.Online topic-aware influence maximization queries[C]//Proceedings of the17th International Conference on Extending Database Technology,Athens,Greece,Mar 24-28,2014:295-306.