DataRETECO’s deduplicated corpus changes what a retrieval ranking can count
The current v1.1 training and development release removes repeated text and maps retired IDs; scores from the earlier corpus need version-aware interpretation.
Model-building research and analysis covering University of Innsbruck.

DataThe current v1.1 training and development release removes repeated text and maps retired IDs; scores from the earlier corpus need version-aware interpretation.