[目的/意义]基于时间网络影响力模型,研究微博话题的时变传播特性。[方法/过程]首先构建微博话题影响力网络模型,给出影响力网络的定义、关键因素分析、模型以及网络权值的计算方法,在此基础上,基于时间网络影响力模型研究微博话题时变传播特性,利用新浪微博平台及DATAMALL的最新微博话题数据仿真分析微博话题随时间的动态传播过程以及对用户的影响力强度。[结果/结论]微博话题随时间的动态传播过程以及对用户的影响力强度之仿真分析结果表明:约有93.3%的话题延迟在1-5小时以内,同时微博话题的影响力网络权值越高,相应的转发评论人数越多,微博话题的影响力也越大。最后将本文提出的TNIM模型与传统的影响力网络模型LDA进行对比,结果显示TNIM模型的影响力网络权值的准确性和稳定性都高于LDA模型,验证了TNIM模型的有效性。
[Purpose/significance] The propagation characteristics micro-blog topic was studied based on the time network influence model.[Method/process] This article first constructed micro-blog topic influential network model, and discussed the definition of the influence of the network, the key factor analysis, model and calculate the network weights.On this basis, the proposed model was based on the propagation time of the network influence(Time Network Influence Model).Using SinaWeibomicro-blog platform and data DATAMALL topic, the simulation of dynamic propagation micro-blog topic with the course of time and users' influence strength was studied.[Result/conclusion] The results showed that:about 93.3% of topics delayed within 5 hours;the higher the micro-blog topic's network influence weight, the more the corresponding number of the forwarding comments, and the greater influence of the micro-blog topic.Finally, the model(TNIM) was compared with the traditional influence of the network model(LDA).The results show the influence of the accuracy and stability of the network weights TNIM model are higher than the LDA model, and the effectiveness of the TNIM model is verified.
[1] 闫幸,常亚平.微博研究综述[J].情报杂志,2011,30(9):61-70.
[2] 赵丽,袁睿翕,管晓宏,等.博客网络中具有突发性的话题传播模型[J].软件学报,2009,20(5):1384-1392.
[3] SREENIVASAN N D,LEE C S,GOH D H L.Tweeting the friendly skies:Investigating information exchange among Twitter users about airlines[J].Electronic library & information systems,2012,46(1):21-42.
[4] WANG DASHUN,SONG CHAOMING,BARABÄI A L.Quantifying long-term scientific impact[J].Science,2013,342(6154):127-132.
[5] BONCHI F,CASTILLO C,IENCO D.Meme ranking to maximize posts virality in microblogging platforms[J].Journal of intelligent information systems,2013,40(2):211-239.
[6] 王晶,朱珂,汪斌强.基于信息数据分析的微博研究综述[J].计算机应用,2012,32(7):2027-2029,2037.
[7] 杨长春,俞克非,叶施仁,等.一种新的中文微博社区博主影响力的评估方法[J].计算机工程与应用,2012,38(25):229-233.
[8] LESKOVEC J,ADAMIC L A,HUBERMAN B A.The dynamics of viral marketing[J].ACM transactions on the Web,2007,1(1):5-9.
[9] PERRA N,GONÇVES B,PASTOR-SATORRAS R,et al.Activity driven modeling of time varying networks[J].Scientific reports,2012,11(2):469-477.
[10] 郭浩,陆余良,王宇,等.基于信息传播的微博用户影响力度量[J].山东大学学报(理学版),2012,47(5):1-6.
[11] 张彦超,刘云,张海峰,等.基于在线社交网络的信息传播模型[J].物理学报,2011,60(5):490-501.
[12] GONZALEZ M C,HIDALGO C A,BARABASI A L.Understanding individual human mobility patterns[J].Nature,2008,453(7196):779-782.
[13] CHENG E,GROSSMAN J W,LIPMAN M J.Time-stamped graphs and their associated influence digraphs[J].Discrete appl.math,2003,128(2/3):317-335.
[14] 王琛,陈庶樵.一种改进的微博用户影响力评价算法[J].信息工程大学学报,2013,14(3):380-384.
[15] MOULAISON H L,BURNS C S.Organization or conversation in Twitter:A case study of chatterboxing[J].Proceedings of the American society for information science and technology,2012,49(1):1-11.
[16] LAMPORT L.Time,clocks,and the ordering of events in a distributed system[J].Communications of the ACM,1978,21(7):558-565.
[17] 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):1956-1964.
[18] NARAYANAM R,NARAHARI Y.A shapley value-based approach to discover influential nodes in social networks[J].IEEE transactions on automation science and engineering,2011,8(1):130-147.
