Skip to content

Enhancing Refugee Management with Data-Driven Intelligence

A new publication in Government Information Quarterly (GIQ) explores how advanced technologies, including large language models, can support refugee management during crises. Titled “Data-driven intelligence in crisis: The case of Ukrainian refugee management,” the paper was authored by Kilian Sprenkamp, Mateusz Dolata, Liudmila Zavalokina, and Gerhard Schwabe. This study explores how large language models and other advanced data-driven technologies can improve refugee support systems, particularly during crises.

The Government Information Quarterly (GIQ) is a leading peer-reviewed journal that focuses on research at the intersection of public administration, technology, and information science, addressing the implications of digital innovation for governance and public service.

The paper addresses the pressing challenge of efficiently managing refugee needs amidst rapidly changing circumstances. Leveraging data from public communication channels, the study introduces R2G (Refugees to Government) – an innovative digital tool powered by large language models, natural language processing and chatbot technologies. This tool allows policymakers to interact dynamically with refugee data, uncovering real-time insights through conversational queries and visual dashboards.

The research paper formulates the design guidelines:

  • Community-Driven Insights: Real-time analysis of refugee communication channels to identify pressing needs directly from the affected communities.
  • Spatial-Temporal Knowledge: Mapping refugee needs based on time and geographic locations to improve targeted responses.
  • Multilingual Data Synthesis: Breaking language barriers using advanced natural language processing techniques to interpret and act upon diverse refugee communications.
  • Interactive Querying via Chatbots: Empowering policymakers with a chatbot interface to explore refugee data intuitively.

The research also outlines policy recommendations for the ethical use of community data, integration of advanced natural language processing techniques in public governance, and fostering a data-driven culture in crisis management.

Read the full publication here to dive deeper into the findings and recommendations.

The ‘Government as a Platform’ project is being funded in the 2nd project call in the innovation programme.