ABSTRACT
Social media platforms are increasingly used during disasters. In the United States, users often consider these platforms to be reliable news sources and they believe first responders will see what they publicly post. While having ways to request help during disasters might save lives, this information is difficult to find because non-relevant content on social media completely overshadows content reflective of who needs help. To resolve this issue, we develop a framework for classifying hurricane-related images that have been human-annotated. Our approach uses transfer learning and classifies each image using the VGG-16 convolutional neural network and multi-layer perceptron classifiers according to the urgency, relevance, and time period, in addition to the presence of damage and relief motifs. We find that our framework not only successfully functions as an accurate method for hurricane-related image classification but also that real-time classification of social media images using a small training set is possible.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Additional information
Funding
Notes on contributors
Matthew Johnson
Matthew Johnson ([email protected]) is a software engineer at Boeing Company. He received a Bachelor of Science in Electrical Engineering (with a focus on data science and software engineering) from the University of Texas at Austin. As a research assistant at the Computational Media Lab, he applied data science to emergency management and meteorological systems using graph-based machine learning models and computer vision.
Dhiraj Murthy
Dhiraj Murthy ([email protected]) is a full professor in the School of Journalism and Media and the Department of Sociology at the University of Texas at Austin, where he also directs the Computational Media Lab. He has authored over 80 articles, book chapters, and papers as well as written the first scholarly book about Twitter (second edition published by Polity Press, 2018). Dr. Murthy’s research explores social media, natural disasters, health, race/ethnicity, and misinformation/disinformation using mixed and computational methods. He is a Co-Editor of the journal Big Data & Society.
Brett W. Robertson
Brett W. Robertson ([email protected]; corresponding author) is an assistant professor in the School of Journalism and Mass Communications at the University of South Carolina. He studies communication technology use in organizational, risk, and mass communication contexts. Dr. Robertson’s research projects explore how individuals use social media and mobile devices in the workplace and disaster-related contexts. Much of his recent focus has been on the communication for disaster preparedness and prevention.
William Roth Smith
William Roth Smith ([email protected]) is an assistant professor in the School of Communication at the University of Tennessee. Dr. Smith’s research focuses on how communication facilitates or constrains organizing processes among loosely structured collectives.
Keri K. Stephens
Keri K. Stephens ([email protected]) is a professor in Organizational Communication Technology and Co-Director of Technology, Information, & Policy Institute in the Moody College of Communication at The University of Texas at Austin. Her research program examines the role of technology in organizational practices and organizing processes, especially in contexts of crisis, disaster, and health. Dr. Stephens has authored over 100 articles appearing in research journals, proceedings, and books. Her two most recent books are the national-level award-winning book New Media in Times of Crisis (2019, Routledge), and the two-time national-level award-winning book Negotiating Control: Organizations and Mobile Communication (2018, Oxford University Press).