Applying Survival Analysis and Count Models to Twitter Data
Document Type
Presentation
Presentation Date
3-26-2018
Abstract or Description
Twitter has a variety of information on it, health topic is one of the popular categories. We used a collection of almost 40,000 tweets extracted from Twitter with #blood pressure from January, 2014 to April, 2015 to investigate the potentially associated factors for popularity (measured by the number of retweet) as well as the survival of tweets (measured by the time frame from the first post to its last retweet). We have found the appearance of a few hashtags significantly decreased the survival of tweets. Furthermore, these hashtags increase( but some decrease) the odds of being retweeted. And other factors significantly associated with the odds include actor's friends count, actor's follower's count, actor's listed count and so on. We explored our results using R, the results do not highlight the potential of hashtag in the application of twitter.
Sponsorship/Conference/Institution
Eastern North American Region International Biometric Society (ENAR)
Location
Atlanta, GA
Recommended Citation
Liu, Congjian, Jingjing Yin, Isaac Chun-Hai Fung, Lindsay A. Mullican.
2018.
"Applying Survival Analysis and Count Models to Twitter Data."
Biostatistics Faculty Presentations.
Presentation 114.
https://digitalcommons.georgiasouthern.edu/biostat-facpres/114
