Despite much research on machine translation (MT) evaluation, there is surprisingly little work that directly measures users' intuitive or emotional preferences regarding different types of MT errors.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback