no code implementations • WS 2020 • Leon Derczynski, Julie Binau, Henri Schulte
We propose two measures for measuring the quality of constructed claims in the FEVER task.