Exploiting Community Emotion for Microblog Event Detection

Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 1159–1168, October 25-29, 2014, Doha, Qatar. c⃝2014 Association for Computational Linguistics Exploiting Community Emotion for Microblog Event Detection Gaoyan Ou1,2, Wei Chen1,2,, Tengjiao Wang1,2, Zhongyu Wei1,3, Binyang Li4, Dongqing Yang1,2 and Kam-Fai Wong1,3 1Key Laboratory of High Confidence Software Technologies, Ministry of Education, China 2School of Electronics Engineering and Computer Science, Peking University, China 3Shenzhen Research Institute, The Chinese University of Hong Kong, China 4Dept. of Information Science & Technology, University of International Relations, China pekingchenwei@pku.edu.cn Abstract Microblog has become a major plat- form for information about real-world events. Automatically discovering real- world events from microblog has attracted the attention of many researchers. Howev- er, most of existing work ignore the impor- tance of emotion information for event de- tection. We argue that people’s emotion- al reactions immediately reflect the occur- ring of real-world events and should be im- portant for event detection. In this study, we focus on the problem of community- related event detection by community e- motions. To address the problem, we pro- pose a novel framework which include the following three key components: mi- croblog emotion classification, community emotion aggregation and community emo- tion burst detection. We evaluate our ap- proach on real microblog data sets. Exper- imental results demonstrate the effective- ness of the proposed framework. 1 Introduction Microblog has become a popular and convenient platform for people to share information about so- cial events in real time. When an external even- t occurs, it will be quickly propagated between microblog users. During propagation process of an event, sufficient amount of users will express their emotions. Taking Sina Weibo1 as an exam- ple, more than 12 percent of users use emoticons2 when reposting an event-related microblog mes- sage. The emotion information can not only help us better understand a given event, but also be u- tilized to discover new events. Figure 1 shows 1http://weibo.com/ 2An icon to indicate user’s emotion, as shown in Table 1. �������� ������� �������� �������� �������� ������� ������� �������� �������� �������� �������� �������� �������� �������� �������� ������� �������� ������� �������� �������� �������� �������� �������� �������� �������� �������� �������� �������� �������� �������� ������� �������� �������� ������� �������� �������� �������� �������� �������� �������� �������� �������� �������� �������� �������� �������� ����� ����� ��� ���� ���� ���� ���� ���� ����� ��� ��� ��� ��� ����� ���� ���� ���� ���� ����� ��� ���� ���� ���� ����� ���� ���� ��� ���� �������� ���� ���� ���� ���� �������� ���� ���� ���� ���� 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Emotion Distribution Day (in March 2011) Happy Anger Sad Fear Figure 1: The global emotion dynamics the emotional distribution dynamics of the over- all microblog messages in March 2011. The sudden change of the public emotion distribu- tion on March 12 indicates a …

Content truncated to 500 words