Annual conference of the European Association for Machine Translation
This conference typically consists of a mix of presentations and poster sessions (presentations of work in a more informal context). This year, it featured four different ‘tracks’:
- a ‘research track’ in which researchers and developers presented the latest trends and developments in MT and translation-related technologies;
- a ‘user track’ in which users from the public sector and industry presented their implementation of and experience with MT systems;
- a ‘project/product track’ which showed advances in MT; and, for the first time,
- a ‘translator’s track’ with short oral contributions by individual translators and a round table to bring MT researchers, developers and vendors closer to the actual users of machine translation.
The keynote speech ‘Human-Centred Translation Technology’ by Sharon O’Brien from Dublin City University was an inspiring talk in which she suggested moving from the notion of ‘computer-aided translation’ to one of translation technology centred on professional translators, ad hoc translators and end users.
One of the main topics of the conference was, as expected, the rapidly developing field of Neural Machine Translation (NMT). A recent advance is Adaptive Neural Machine Translation where the post edited output produced by translators is used to improve the machine translation engines. This means that translators do not have to keep correcting the same MT mistakes over and over again.
The proceedings of the conference are available at http://eamt2018.dlsi.ua.es/proceedings-eamt2018.pdf