TALN⊕RÉCITAL 2022 — Avignon
27 juin ⇢ 1er juillet 2022
La conférence intègrera l’intervention de deux conférencières invitées, le mardi et le jeudi.
A multilingual is someone who speaks more than one language for communication in his/her daily life. Multilingualism is commonly practiced at the individual or societal levels in most parts of the world (e.g., Europe, Asia, Africa). However, current language technologies are usually built around monolingual assumptions which do not always reflect the linguistic variation in real world communication. The first goal of this talk is to familiarize the computational linguistics community with the key concepts around multilingualism. Secondly, I will discuss the challenges and opportunities for building linguistically diverse, inclusive and multilingual language technologies through giving examples from high and low resource languages and its speakers.
A. Seza Doğruöz is a faculty at Ghent University. Her research focuses on analyzing multilingual language use and linguistic variation between humans and developing language technologies accordingly. Her publications cover the themes related to Computational Sociolinguistics, Multilingualism & Linguistic Variation, Curating Language Resources (especially for low resource languages) and Open Domain Dialog Systems. Examples of her recent publications are:
Language technology underpins many of the applications and platforms that enable our digitally enhanced lives (virtual assistants, search engines, translation tools, spell-checkers, language learning tools, etc.). Yet these advances do not benefit all languages equally. Due to a lack of sufficient linguistic support, users often need to revert to using another language. Such a language shift plays a major role in the risk of digital extinction, i.e. an eventual decline in language use due to lack of technological support.
This talk focuses on Irish — an official EU language, and considered a low-resourced language in terms of digital support. ‘Low-resourced’ not only means that there is a severe lack of speech and language applications available for Irish speakers to use, but it also means that the fundamental tools and language resources required to build these technologies are also lacking. Irish is also in a precarious position while it competes alongside the most technologically supported language in the world — English. I will talk about our work at Dublin City University where we are taking some steps towards addressing this risk of digital extinction through the development of NLP tools and resources for Irish.
Dr. Teresa Lynn is a Research Fellow at the ADAPT Centre in Dublin City University. She was awarded her PhD in 2015 from both DCU and Macquarie University Sydney, and was a recipient of the 2014-2015 Enterprise Ireland Fulbright Award, carrying out research at Saint Louis University, Missouri, USA.
Her main research interests lie in developing tools and resources for Irish language technology. She is the principal investigator on the GaelTech project, funded by the Irish Government Department of the Gaeltacht, which covers various research topics in Irish NLP. She is also a core member of the European Language Equality project and Ireland’s National Anchor Point for the ELRC (European Language Resource Coordination), overseeing national data collection for Irish machine translation. Other projects include ELRI (European Language Resource Infrastructure), the National Digital Plan for Irish and the Universal Dependencies Project. Her research covers treebank development, syntactic parsing, social media NLP and multiword expressions. Teresa also worked in industry for several years, namely in the areas of localisation, NLP and machine translation.