Start Date

12-13-2015

Description

Social media enable users to express their emotion promptly, helping health policy makers to gauge public sentiment of disease outbreak. In this research, we developed an approach to social-media-based public health informatics and built a proof-of-concept system named eMood that helps to collect, analyze, and visualize Ebola outbreak discussions on Twitter. Our approach uses a comprehensive lexicon to identify emotion categories and present analysis findings of users’ network relationship and influence patterns. We compared two methods of identifying user influence, user centrality and emotion entrainment, by using 255,118 tweets posted by 210,900 users in January 2015. Empirical results show that both methods identified highly influential users. Regression analysis of user influence rank and emotion scores demonstrates significant relationship between user influence and each emotion category. These results should provide strong implication for understanding social actions and for collecting social intelligence for public health informatics.

Share

COinS
 
Dec 13th, 12:00 AM

eMood: Modeling Emotion for Social Media Analytics on Ebola Disease Outbreak

Social media enable users to express their emotion promptly, helping health policy makers to gauge public sentiment of disease outbreak. In this research, we developed an approach to social-media-based public health informatics and built a proof-of-concept system named eMood that helps to collect, analyze, and visualize Ebola outbreak discussions on Twitter. Our approach uses a comprehensive lexicon to identify emotion categories and present analysis findings of users’ network relationship and influence patterns. We compared two methods of identifying user influence, user centrality and emotion entrainment, by using 255,118 tweets posted by 210,900 users in January 2015. Empirical results show that both methods identified highly influential users. Regression analysis of user influence rank and emotion scores demonstrates significant relationship between user influence and each emotion category. These results should provide strong implication for understanding social actions and for collecting social intelligence for public health informatics.