Date and time: 11/12/25, 5pm CET
Speaker: Antonio Toral Ruiz, Department of Informatics and Languages, University of Alicante
Presenter: Elena Lloret, University of Alicante
Abstract:
Machine Translation (MT), like other NLP applications, has experienced a spectacular improvement over the last decade. This progress was initially driven by encoder-decoder neural architectures and, more recently, by the advent of Large Language Models (LLMs). Indeed, according to evaluations from the latest editions of WMT—the principal competition for this task—MT is thought to have already reached the level of professional human translation for various text types across multiple language pairs.
But, is this really the case? Can we consider the task of MT concluded? In this presentation, I will argue that MT can and should be evaluated beyond the mere presence of errors. Adopting concepts and methodologies from the discipline of Translation Studies, I will present a research line focused both on the generation of more creative and natural machine translations, and on the development of evaluation methodologies to measure these two characteristics. This research is framed as a use case of interdisciplinarity applied to NLP, with the goal of addressing the limits of current technology.
Bio: Antonio Toral works as a Distinguished Professor at the Universitat d'Alacant (Spain). Previously, he was an Associate Professor at the University of Groningen (Netherlands), where he coordinated the Computational Linguistics research group, which had over 40 members.
His research interests include the application of Machine Translation (MT) to literary texts, MT for low-resource languages, and the quantitative analysis of machine and human translations. He coordinated the Abu-MaTran project, which was highlighted by the European Commission as a success story, won the Best Paper Award at the 2019 MT Summit conference for his work on post-editing, and co-founded the Spanish Linguistics Olympiad.
He completed his postdoctoral research at Dublin City University (Ireland), and his doctoral studies at the Universitat d'Alacant and the Istituto di Linguistica Computazionale (Italy).
Registration (mandatory): https://zoom.us/webinar/register/WN_ywhkexFHS4KXqIIOjTuRAw
Link to the talk:
