id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_rzcndqwjcbdidbg5nsppqfii2q	André F. T. Martins	Pushing the Limits of Translation Quality Estimation	2017	14	.pdf	application/pdf	8265	805	68	neural model into a rich feature-based wordlevel quality estimation system. the output of an automatic post-editing system as an extra feature, obtaining striking results on WMT16: a word-level F MULT1 score of 2014; Kim and Lee, 2016), and systems that combine linear and neural models (Kreutzer et al., 2015; Figure 1: Example from the WMT16 word-level QE training set. translation (MT), its manual post-edition (PE), and the conversion to word quality labels made with the TERCOM tool predict word-level quality labels (yielding APEQE, the datasets above, the word quality labels are obtained automatically by aligning the translated and baseline systems provided in WMT15/16, we include features that depend on the target word and Table 11: Performance of the several word-level QE systems on the WMT16 development and test datasets. Table 11: Performance of the several word-level QE systems on the WMT16 development and test datasets. the WMT16 Word-Level Translation Quality Estimation Shared Task.	./cache/work_rzcndqwjcbdidbg5nsppqfii2q.pdf	./txt/work_rzcndqwjcbdidbg5nsppqfii2q.txt