In this session, we will give an account of the harmonization process subsidized by the European Commission as part of the QT21 project, followed by two use cases. The aim of QT21 was to develop translation models for challenging languages and resource scenarios, to produce improved evaluation and continuous learning from mistakes guided by a systematic analysis of quality barriers and informed by human translators. The resulting error-typology DQF-MQM is integrated in several translation technologies resulting in adoption by major translation buyers and vendors and is becoming a game changer in translation quality evaluation in our industry. Two companies, Tableau Software and Dell, will describe how they have adopted the DQF-MQM standard in their workflows.
Normalizing Quality Management — Two Use Cases with DQF