Abstract:

We present a prescriptive recommendation framework for AI-based solutions for managing false information on social media during disasters. This framework is built using a design science research approach. Three main considerations were made: (1) the type of disaster; (2) whether an adaptive AI model is needed; and (3) finding an optimal balance between AI technologies including cloud-based, federated learning, and Edge AI. The framework can help disaster managers and emergency-response organizations select AI-based solutions to counter the spread of false information on social media.