Cross-lingual social sensing for disaster monitoring: A pan-European framework for cross-border information flow analysis and antifragility indicators

Ghoroghi, Ali, Hodorog, Andrei ORCID: https://orcid.org/0000-0002-4701-5643, Rezgui, Yacine ORCID: https://orcid.org/0000-0002-5711-8400, Ghaemi, Afrouz and De Nardi, Cristina 2026. Cross-lingual social sensing for disaster monitoring: A pan-European framework for cross-border information flow analysis and antifragility indicators. Progress in Disaster Science 32 , 100715. 10.1016/j.pdisas.2026.100715

PDF - Published Version Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (1MB)

Abstract

Disasters routinely propagate across administrative boundaries and linguistic communities, yet social sensing instruments for monitoring them remain largely monolingual and single-city. This paper presents a cross-lingual social sensing framework for disaster and disruption monitoring at pan-European scale. A supervised multilingual classifier was applied to 13,388 social media posts from 23 countries in 10 languages, categorising disruption discourse into 11 event types with held-out accuracy of approximately 0.83 (Cohen’s κ≈0.79). Posts were geographically normalised across five pilot Areas of Interest, and a geolinguistic information-flow network was constructed to characterise cross-border disaster communication pathways. Five pilot antifragility indicators were defined with baseline measurement architectures. Transport disruptions constituted the dominant signal (23.8% of classified discourse). Sentiment was predominantly neutral (94.5%), suggesting urgency-based classification affords greater discrimination than polarity for monitoring applications. Cross-border information flows clustered along geographic and linguistic boundaries whilst sustaining bridging channels carrying disaster signals across jurisdictions. The framework contributes a replicable cross-lingual sensing pipeline for disaster monitoring, a cross-border communication analysis, and a pilot indicator layer for antifragility-oriented disaster system assessment, aligned with Sendai Framework Priorities 1 and 4.

Item Type:	Article
Date Type:	Publication
Status:	Published
Schools:	Schools > Engineering
Publisher:	Elsevier BV
ISSN:	2590-0617
Date of First Compliant Deposit:	21 September 2026
Date of Acceptance:	7 September 2026
Last Modified:	21 Sep 2026 16:00
URI:	https://orca.cardiff.ac.uk/id/eprint/189748

Actions (repository staff only)

Edit Item