From the 10-12 May 2017, representatives from the Centre attended the International Annual Meeting on Computer-Assisted Translation and Terminology (JIAMCATT) hosted at the International Criminal Court in The Hague.
JIAMCATT is an annual event that provides its partners with a forum for debate, exchange of expertise and cooperation in the fields of computer-assisted terminology and translation, interpretation and documentation retrieval. It provides an ideal platform for exchange with the research community, by bringing together more than 160 language professionals representing European institutions, international organisations and the member states.
For this year’s edition, the organisers had chosen the following theme: The power of simplicity: practical solutions for today’s real-world problems. In an increasingly complex business and technological environment, the quest for simplicity is the common motivator. There is much to be gained by reviewing and modernising processes, tools and working methods.
During the meeting, there were workshops on several research areas, one of which was authoring. Authoring refers to the process of defining the structure of original texts being sent for translation by identifying their components and reusing them during each new reproduction of an original text. This has caught the growing attention of the main actors in the translation process.
The European Commission presented recent developments in the evolution from the MT@EC machine translation system to eTranslation within the Connecting Europe Facility (CEF) framework. This included steps being taken to automatically harvest feedback from translators, and a short description of new engines for morphologically complex languages based on Neural Machine Translation technology. At JIAMCATT 2017, it was made crystal clear that the translation community can no longer postpone working on the integration of Statistical Machine Translation (SMT) / Neural Machine Translation (NMT) engines for the most productive languages. SMT/NMT engines are not only integrated in Computer Assisted Translation solutions for translators/revisers and used during the processing phases of a translation management system, but can also serve as standalone components connected to online service solutions. JIAMCATT recognises that the emerging neural machine translation paradigm opens up new business opportunities.
The Translation Centre took the opportunity of this year’s event to share its experience in the field of subtitling. The quality-based workflow implemented at the Centre was of key interest to many participants who were also able to take part in an afternoon of demo sessions. EU institutions and agencies are increasingly interested in posting short videos on social media and subtitling them to reach EU citizens in their own languages. This explains why more and more EU agencies are turning to the Translation Centre for this new language service, which requires different skills, specialised software, as well as a specific workflow.