Predicting Tweet Retweetability during Hurricane Disasters

Predicting Tweet Retweetability during Hurricane Disasters

Venkata Kishore Neppalli (University of North Texas, Computer Science and Engineering, Denton, TX, USA), Cornelia Caragea (University of North Texas, Computer Science and Engineering, Denton, TX, USA), Doina Caragea (Kansas State University, Department of Computer Science, Manhattan, KS, USA), Murilo Cerqueira Medeiros (University of North Texas, Computer Science and Engineering, Denton, TX, USA), Andrea H. Tapia (Pennsylvania State University, University Park, PA, USA) and Shane E. Halse (Pennsylvania State University, University Park, PA, USA)
DOI: 10.4018/IJISCRAM.2016070103
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Twitter is a vital source for obtaining information, especially during events such as natural disasters. Users can spread information on Twitter either by crafting new posts, which are called “tweets,” or by using the retweet mechanism to re-post previously created tweets. During natural disasters, identifying how likely a tweet is to be retweeted is crucial since it can help promote the spread of useful information in a social network such as Twitter, as well as it can help stop the spread of misinformation when corroborated with approaches that identify rumors and misinformation. In this paper, we present an analysis of retweeted tweets from two different hurricane disasters, to identify factors that affect retweetability. We then use these factors to extract features from tweets' content and user account information in order to develop models that automatically predict the retweetability of a tweet. The results of our experiments on Sandy and Patricia Hurricanes show the effectiveness of our features.
Article Preview

Introduction

In response to increased online public engagement and the emergence of digital volunteers, emergency responders have sought to better understand how they too can use online media to communicate with the public and collect intelligence (Denef, Bayerl, & Kaptein 2013; Latonero & Shklovski 2011; Sutton et al. 2014; St. Denis, Palen, & Anderson 2014). Many emergency decision makers see the data produced through crowdsourcing as ubiquitous, rapid and accessible - with the potential to contribute to situational awareness (Vieweg et al. 2010). As the use of public social media in crisis increased, emergency responders started to take notice of the way citizens engaged in social media and the information exchanges that took place there (Hughes & Tapia, 2015). Consequently, responders began to consider if social media might be a useful tool for their practice. Research revealed that social media could be used to distribute information quickly to a wide-spread audience (Kodrich & Laituri 2011) and to engage more directly in a two-way conversation with members of the public (Hughes & Palen 2012). However, incorporating the products of digital volunteer activity into professional emergency practice has proven to be challenging due to issues with credibility, liability, training, and organizational process and procedure (Hughes & Palen 2012; Tapia et al. 2011).

The information that the public produced looked to be useful, as researchers showed that it could contribute to situational awareness during a crisis event (Cameron et al. 2012; Ireson 2009). According to Starbird, Munzy and Palen (2012), social media data that can be identified as coming from local bystanders to a disaster can be extremely important to emergency responders. Most of the social media data surrounding a disaster are derivative in nature: information in the form of reposts or pointers to information available elsewhere (Starbird et al., 2010). These derivative data are abundant, as a form of noise that must be filtered out to arrive at the signal of good data (Anderson & Schram, 2011). A small subset of the data comes from locally affected populations in the form of citizen reports (Starbird et al., 2010). Starbird et al. (2010) assert that bystanders “on the ground are uniquely positioned to share information that may not yet be available elsewhere in the information space … and may have knowledge about geographic or cultural features of the affected area that could be useful to those responding from outside the area.”

Complete Article List

Search this Journal:
Reset
Open Access Articles
Volume 9: 4 Issues (2017): 1 Released, 3 Forthcoming
Volume 8: 4 Issues (2016)
Volume 7: 4 Issues (2015)
Volume 6: 4 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing