Automating Translation
Sobre aquest llibre
This book, authored by leading experts, demystifies machine translation, explaining its origins, its training data, how neural machine translation and LLMs work, how to measure their quality, how translators interact with contemporary systems for automating translation, and how readers can build their own machine translation or LLM. In later chapters, the scope of the book expands to look more broadly at translation automation in audiovisual translation and localisation. Importantly, the book also examines the sociotechnical context, focusing on ethics and sustainability. Enhanced with activities, further reading and resource links, including online support material on the Routledge Translation studies portal, this is an essential textbook for students of translation studies, trainee and practising translators, and users of MT and multilingual LLMs.
Additional resources are available on the Routledge Translation Studies Portal.
Sobre l'autor
Joss Moorkens is Associate Professor at the School of Applied Language and Intercultural Studies and Science Lead at the ADAPT Centre at Dublin City University, Ireland. He is General Co- Editor of the journal Translation Spaces, author and editor of several books, articles, chapters, and special issues on translation technology, and sits on the board of the European Masters in Translation network.
Andy Way is Professor of Computing and Co- Founder of the ADAPT Centre at Dublin City University, Ireland. He was previously editor of the Machine Translation journal for 15 years, and president of both the European and International Associations for Machine Translation. He has over 450 publications, including five books on machine translation.
Séamus Lankford is a Computer Science lecturer with over 25 years’ experience at the Munster Technological University, Ireland. He has published extensively on the topic of machine translation. The focus of his doctoral thesis was the enhancement of NMT of low- resource languages through corpus development, human evaluation, and explainable AI architectures.
