Standard Set

We present results of experiments designed to assess the usefulness of a new technique for the evaluation of translation quality, comparing human rankings with automatic measures. The basis of our …...