Error Span Annotation: A Balanced Approach for Human Evaluation of Machine Translation

Open in new window