disruption_v1: An audited database of operational disruptions on the China–Europe Railway Express (2014–2026)
收藏资源简介:
disruption_v1 records 64 operational disruption events on the China–Europe Railway Express (CRE), with a focus on the northern corridor: border closures, capacity overflows, geopolitical conflicts and sanctions, COVID-era interruptions, infrastructure failures, regulatory changes, strikes and weather events between January 2014 and February 2026 (59 events fall within the 2018–2025 GDELT signal window used by the companion studies). Events were extracted from 236 industry news articles (RailFreight.com, Yicai Global, and others) with LLM-assisted extraction followed by human audit; each event carries a severity level, affected corridor nodes aligned to a 41-node network, multi-source verification flags, quoted evidence snippets and source URLs. Version 1.1 adds audit-verified omissions discovered during the false-alarm audit of the companion calibration study (currently the Kuźnica border closure of 2021-11-09). The deposit also includes three deep-dive case corpora of LLM-extracted, human-audited sub-events for structurally distinct geopolitical shocks: G.1 — the Russia–Ukraine war onset, 2022 (war-plus-sanctions background; 67 sub-events); G.2 — Russian Federation Directive No. 1374 export controls, 2024 (gradual policy event; 48 sub-events); and G.3 — the Poland–Belarus border closure, September 2025 (sudden military event; 59 sub-events). Files: the event database (parquet and CSV), the v1.1 additions, metadata of the 236 source articles (no full article texts are redistributed), the 41-node corridor table, and the three case corpora. A complete data dictionary is provided in README.md inside the archive. Quoted evidence snippets are short excerpts from publicly available news reporting, reproduced under quotation for research documentation. This database serves as the shared ground truth of two companion research articles, which will be added to this record's metadata as they are published.



