A Comprehensive Study of Disaster Support Mobile Apps

Abstract

Context: Disasters are a common global occurrence with climate change leading to increase both their frequency and intensity. To reduce the impact of these disasters on lives and livelihoods it is important to provide accurate warnings and information about recovery and mitigation. Today most emergency management agencies deliver this information via mobile apps.Objective: There is a large collection of disaster mobile apps available across the globe. But a detailed study is not yet conducted on these apps and their reviews to understand their key features and user feedback. In this paper we present a comprehensive analysis to address this research gap.Method: We conducted a detailed analysis of 45 disaster apps and 28,161 reviews on these apps. We manually analysed the features of these 45 apps and for review analysis employed topic modelling and sentiment analysis techniques.Results: We identified 13 key features in these apps and categorised them in to the 4 stages of disaster life cycle. Our analysis revealed 22 topics with highest discussions being on apps alert functionality, app satisfaction and use of maps. Sentiment analysis of reviews showed that while 22\% of users provided positive feedback, 9.5\% were negative and 6.8\% were neutral. It also showed that signup/signin issues, network issues and app configuration issues were the most frustrating to users. These impacted user safety as these prevented them from accessing the app when it mattered most.  Conclusions: We provide a set of practical recommendations for future disaster app developers. Our findings will help emergency agencies develop better disaster apps by ensuring key features are supported in their apps, by understanding commonly discussed user issues. This will help to improve the disaster app eco-system and lead to more user friendly and supportive disaster support apps in the future.

Keywords: mobile apps, disaster apps, crisis management, user reviews, analysis, recommendations