摘要
【目的/意义】将时间序列分析方法引入情感分析,可以对微博突发事件衍生舆情作出科学预测,为政府掌握舆情情感走势,从而根据舆情发生的不同阶段采取相应的导控策略提供合理的意见与指导。【方法/过程】结合突发事件衍生舆情的特点,采用词集合并法、SO-PMI、PMI-IR等方法构建了微博突发事件衍生舆情专属情感词典,随后利用该情感词典和时间序列分析方法对"6.22"杭州保姆纵火案衍生舆情事件进行实证分析。【结果/结论】对该事件的日均情感值进行计算,与实际情感值拟合程度较好,证明了建立的衍生舆情情感词典及时间序列模型较为科学,可以为政府选择相应的策略及应对时机提供一定的参考。
【Purpose/significance】The time series analysis method is integrated into the sentiment analysis, which can make scientific prediction to the derivative public opinion of microblog, and then provide the reasonable advice and guidance for the government based on the stage of the public opinion.【Method/process】Combined with the characteristics of the derived public opinion for microblog emergencies, this paper uses the method of word set merging, SO-PMI and PMIIR to construct the sentiment lexicon suitable for the special sentiment analysis, then the Dictionary and time series analysis method are used to analyze the case of "6.22" Hangzhou nanny arson.【Result/conclusion】The average daily emotion value of this event is calculated and the sentiment value get good fitting degree, which proves that the established dictionary is affective and the time series model of the derived public opinion is accurate. So it can provide some reference for the government to choose the coping strategy and the time to deal with it.
引文
1中国互联网络信息中心.第41次《中国互联网络发展状况统计报告》[EB/OL].http://www.cnnic.net.cn/hlwfzyj/hlwxzbg/hlwtjbg/201803/t20180305_70249.htm,2018-03-01.
2刘亚男.我国网络舆情研究现状述评[J].情报杂志,2017(5):95-100.
3刘毅.网络舆情研究概论[M].天津:天津人民出版社2007.
4 Bing Liu.Sentiment Analysis:Mining Opinions,Sentiments,and Emotions[M].北京:机械工业出版社,2017.
5 Hu,Xia,Lei Tang,Jiliang Tang,and Huan Liu.Exploiting Social Relations for Sentiment Analysis in Microblogging[C]//In proceeding of the ACM International Conference on Web Seatch and Data Mining(WSDM-2013).2013.
6 Zhou Y,Moy P.Parsing Framing Processes:The Interplay Between Online Public Opinion and Media Coverage[J].Journal of Communication,2007,57(1):79-98.
7 The influence of local majority opinions on the dynamics of the Sznajd model[J].Nuno Crokidakis.Journal of Physics:Conference Series,2014,(487):1-7.
8王澍贤,陈福集.意见领袖参与下微博舆情演化的三方博弈分析[J].图书馆学研究,2016,(1):19-25.
9刘翠娟,刘箴,柴艳杰,等.基于微博文本数据分析的社会群体情感可视计算方法研究[J]..北京大学学报(自然科学版),2016,(1):178-186.
10温淑春.网络舆情对政府管理的影响及其应对机制探讨[J].理论与现代化,2009,(5):103-107.
11 Noelle-Neumann,E.Zhe Spiral of Silence:Public OpinionOur Social Skin(Second edition)[M].Chicago:the university of Chicago Press,1993:59.
12 Zhou Y,Moy P.Parsing Framing Processes:The Interplay Between Online Public Opinion and Media Coverage[J].Journal of Communication,2007,57(1):79-98.
13 Dave,Kushal,Steve Lawrence,and David M.Pennock.Mining the Peanut Gallery:Opinion Extraction and Semantic Classification of Product Reviews[C]//In Proceedings of International Conference on World Wide Web.2003
14 Chen,Bi,Leilei Zhu,Daniel Kifer,and Dongwon Lee.What Is an Opinion About?Exploring Political Standpoints Using Opinion Scoring Model[C]//In proceedings of AAAI Conference on Artificial Intelligence,2010.
15董靖巍.基于复杂网络的网络舆情动态演进影响机制研究[D].哈尔滨:哈尔滨工业大学,2016.
16赵妍妍,秦兵,刘挺.文本情感分析[J].软件学报,2010,21(8):834-848.
17聂淑媛.时间序列分析的历史发展[J].广西民族大学学报(自然科学版),2012,(1):24-28.
18周咏梅,阳爱民,林江豪.中文微博情感词典[J].山东大学学报(工学版),2014,44(3):36-40.
19 Turney P D.Mining the Web for Synonyms:PMI-IR Versus LSA on TOEFL[C]//In Proceedings of the 12th European Conference on Machine Learning,Freiburg,Germany,2001:491-502.
20 Turney P D.Mining the Web for Synonyms:PMI-IR Versus LSA on TOEFL[C]//In Proceedings of the 12th European Conference on Machine Learning,Freiburg,Germany,2001:491-502.
21 Turney P D,Littman M L.Unsupervised Learning of Semantic Orientation from a Hundred-Billion-Word Corpus[R].National Research Council of Canada,Tech.Rep:EGB1094,2002.