Alert notifications for governmental disaster response via instant messaging applications

Abstract

Disaster response decision-making requires adaptation to increasing information overflow, namely the continuing growth of the amount and sources of data. Conversational agents, software applications interacting with users conversationally, have been adopted to enhance accessibility to information for this purpose. This study conceptualizes the design and implementation of alert notifications via instant-messaging-based conversational agents for governmental disaster response to achieve timely and accurate awareness to support effective decision-making. The data inventory scheme to support comprehensive implementation and the Rich Notification Architecture (RNA) for alert notifications are constituted. The data inventory scheme distinguishes data providers for service availability. The RNA improves service availability and information accessibility by dissecting complex procedures to construct an alert notification from a data source. The approach is adopted in two central governmental agencies in Taiwan, demonstrating its applicability to different disaster response goals.