Dear all, This year we have exciting new data for the task, coming from Wikipedia articles. For the first time, we provide sentence-level annotation based on Direct Assessment, as well contextual information about the sentences (the article where they came from), the NMT models and the scores they produce. This data is available for 7 language pairs, with 8K training/dev samples per language pair. In addition, we have the traditional tasks with HTER and MQM-based annotation for sentence, word and document-level QE.
Please check the task website: http://www.statmt.org/wmt20/quality-estimation-task.html, where you can find both the training and test datasets and links for the submission systems. -- Lucia Specia (on behalf of task organisers) www.dcs.shef.ac.uk/~lucia/
_______________________________________________ Mt-list site list Mt-list@eamt.org http://lists.eamt.org/mailman/listinfo/mt-list