Geolocating Social Media for Crisis Management: A Systematic Literature Review

Abstract

To assess the state of social media geolocation research and inform its application to crisis management, we conducted a systematic literature review of 46 peer-reviewed studies published between 2020–2024. These studies establish a pre-generative AI baseline for understanding how emerging large language model (LLM) approaches may extend geolocation capabilities. Our review reveals inconsistently defined geolocation inference goals and spatial resolution targets, a predominant—though not universal—focus on Twitter/X, and goal-contingent uses of methods and data sources that constrain operational applicability in crisis contexts. Synthesizing these findings, we introduce a standardized conceptual framework for classifying geolocation inference goals and spatial resolutions, enabling consistent comparison across existing and LLM-based approaches. We conclude by highlighting the need and opportunities to align social media geolocation research with operational needs and requirements in crisis management contexts.