Uncovering Hidden Behavioral Patterns in the Era of “We Media”: Modeling Spatio-Temporal Dynamics for Twitter News
Document Type
Presentation
Publication Date
2017
Abstract
This research presents a Bayesian statistical model to examine spatio-temporal effects for Twitter use when reporting important events or news. The proposed model tests the Twitter News data surrounding the United States Supreme Court’s Myriad Genetics, Inc. June 13, 2013 decision and its impact on direct-to-consumer genetic testing and gene patenting. The model demonstrates the sensitivity in distinguishing the behaviours of Twitter users’ followers with and without adjusting spatio-temporal effects. It was also found that media professionals’ tweets were coming thick and quick, and producing “waves” of engagement of followers. However, grassroots actively participate in tweeting and constantly engage more followers. The model maps tweets across the spatial heterogeneity and temporal evolution in the early and post recognition and discussion of events. These findings demonstrate the importance of spatio-temporal effects to influence professionals or non-professionals for tweeting. The model also guided researchers to detect sub-events with low latency.
Was this content written or created while at USF?
Yes
Citation / Publisher Attribution
Presented at the 2017 iConference on March 22-25, 2017 in Wuhan, China
Scholar Commons Citation
Huang, Hong; Yu, Han; Andrews, James E.; Yoon, JungWon; and Burgess, Kelsey L., "Uncovering Hidden Behavioral Patterns in the Era of “We Media”: Modeling Spatio-Temporal Dynamics for Twitter News" (2017). School of Information Faculty Publications. 427.
https://digitalcommons.usf.edu/si_facpub/427
