Humanitarian data infrastructures for missing migrants: A multimodal and ethics-integrated framework

<ns3:p>The disappearance of migrants is a global humanitarian crisis aggravated by structural data deficiencies across fragmented migration routes. This study adopts a sociotechnical systems perspective to develop an ethics-integrated infrastructure designed to convert multimodal digital traces into actionable intelligence for search, identification, and policy accountability. We propose a novel typology classifying 32 trace types across ten domains ranging from geospatial markers to psychosocial signals like emoji usage, facilitating the systematic analysis of decentralized digital evidence. This classification underpins a modular architecture comprising five interconnected cores: ingestion, preprocessing, evidence integration, AI, and analysis. The system enables entity resolution and spatiotemporal reconstruction while incorporating provenance tracking and explainable AI to deliver early warnings and OCHA-aligned decision support. By embedding stringent cross-layer safeguards including dynamic consent, auditability, and interoperability, the framework ensures adherence to FAIR data principles and international human rights standards. This research offers a scalable, replicable method for validating digital evidence across the migration cycle, providing a bridge between computational innovation and rights-based migration governance. Future work targets operational pilots to assess feasibility and impact on triage efficiency, identification accuracy, and family notification protocols.</ns3:p>