Abstract
The use of social media today is not only ubiquitous and an integral part of everyday life but is also increasingly relevant before, during, or after emergencies. Data produced in these contexts, such as situational updates and multimedia content, is disseminated across different social media platforms and can be leveraged by various actors, including emergency services or volunteer communities. However, the dissemination of several thousand or even millions of messages during large-scale emergencies confronts analysts with challenges of information quality and overload. Hence, crisis informatics as a research domain seeks to explore and develop systems that support the collection, analysis, and dissemination of valuable social media information in emergencies. This chapter presents the social media API (SMA), which is a multi-platform service for gathering big social data across different social media channels and analyzing the credibility and relevance of collected data by the means of machine learning models. Based on the lessons learned from both the implementation process and user-centered evaluations in multiple emergency settings, this chapter discusses core challenges and potentials of the SMA and similar services, focusing on (1) the multi-platform gathering and management of data, (2) the mitigation of information overload by relevance assessment and message grouping, (3) the assessment of credibility and information quality, and (4) user-centered tailorability and adjustable data operations.KeywordsBig social dataSocial mediaCrisis informaticsInformation refinementMulti-platform services
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call