The pace of progress in the field of language technology is faster than ever. Less than ten years since the generalized deployment of neural machine translation (NMT), we are now witnessing an even more significant breakthrough with the advent of large language models (LLMs). Availability of these tools has already changed the way we translate and teach translation. In the coming years they are likely to have a radical impact on every aspect of our lives.
An understanding of how NMT and LLMs work is essential to take full advantage of them and to understand their limitations, both in general and with respect to each other. For this reason, in this webinar we describe the essential principles behind NMT and LLMs, and how they compare to each other when it comes to translating. We approach these topics from the perspective of professional users (translators, students, researchers and teachers of translation) and focus in particular on issues related to language data (re)use. More specifically, we shall cover the following:
- Training
- Adaptation/stylised translation
- Quality comparisons
- Data privacy
- Copyright
- Synthetic data
- Sustainability
- Ease of access
The webinar will be in English.
Teachers
- Silvia Bernardini (Laurea – Bologna, MPHil – Cantab, PhD – MDX) is professor of English linguistics and translation studies at the Department of Interpreting and Translation of the University of Bologna, where she teaches translation from English into Italian and corpus linguistics. She has published widely on corpus use in translator education and for translation practice and research. Her research interests also include the investigation of the points of contact between translation and interpreting and translation and non-native writing seen as instances of bilingual language use in language contact situations, and the challenges faced by the language professions at a time of fast technological and societal changes.
- Claudia Lecci graduated in 2007 in Specialized translation and translation for the publishing industry at the Advanced School of Modern Languages for Interpreters and Translators (now Department of Interpreting and Translation – DIT), Forlì, University of Bologna. She currently teaches Methods and Technologies for Interpreting, CAT Tools and Post-editing at DIT, where she is also a tutor in the field of quality improvement in computer-assisted translation and terminology training. She works as a consultant for language service providers and companies in the field of translation and interpreting technologies. She is a Trados Authorized Trainer for Trados Studio and MultiTerm.
- Federico Garcea is a computer scientist who graduated in 1998 in Operation Research at the University of Bologna. He previously worked for Microsoft Research, where he directed the development of Machine Translation engines and services. He currently teaches Machine Translation at Department of Interpreting and Translation – DIT, Forlì, University of Bologna. His current research focuses on translation of artistic and cultural content, as well as preserving intangible cultural heritage through technological innovations. He also works on assessment of impact of translation technologies, with the aim of promoting and improving user interaction and participation in services and content, breaking down language barriers.
Date and time
Friday, May 31st, 13:00-14:00
Registration
The webinar is free, you only need to fill out this form and we’ll send you the link the day before the event: https://forms.office.com/e/F0geztHsFS
Registration closes on May 29.