thomas gaillat
Email:
I am a Postdoctoral Researcher and Deputy Unit leader at Insight Centre for Data Analytics National University of Ireland Galway. My research focus is Computational Linguistics grounded in corpus analysis and NLP.
Before joining Insight, I did my PhD in Linguistics at the University of Sorbonne Paris Cité. My PhD focused on corpus interoperability and referential issues related to the use of deictics in the speech of learners of English. I was also an English teacher for ten years at the University of Rennes in France and deployed numerous CALL systems. I also lectured on CALL course design. More recently, I have lead research work in the fields of Computational Semantics and Machine Learning for Sentiment Analysis as part of the SSIX project. My research domains include learner and native corpus annotation and interoperability, reference, subjectivity and Machine Learning methods for linguistic analysis.
Lay summary of my research:
My research focuses on using AI tools to understand the linguistic processes involved in the production of language. I am particularly interested in understanding how people learn a language. It is generally admitted that learners of a second language go through universal stages in their progression. I focus on how learners make use of specific language systems such as making references to items in speech. I want to see what differences there are between learners and native speakers. I also try to discover if learners of different native tongues go through the same stages. To conduct my research I use collections of texts/transcripts, called corpora, produced by learners and natives. These corpora can be processed by using AI technologies to analyse the words and contexts produced by learners and to measure the differences with native speakers.
The expected outcomes of this research are twofold. Firstly, by looking into the way AI tools make decisions, it is possible to better understand the linguistic criteria that are significant in language production. Secondly, this knowledge can be used to create AI systems dedicated to the real-time analysis of learner essays, e.g., language level identification, error detection and recommendation systems etc.
Publications
- Conference Name: Annual Meeting of the Association for Computational Linguistics) ,Authors
Thomas Gaillat, Bernardo Stearns, Ross McDermott, Gopal Sridhar, Manel Zarrouk, Brian Davis
Publication DetailsType:Proceedings: Proceedings of the First Workshop on Economics and Natural Language Processing, Date: 2018Location: Australia,Refereed Conf./Meeting Proceeding - Conference Name: The Language Resources and Evaluation Conference ,Authors
Thomas Gaillat, Manel Zarrouk, André Freitas, Brian Davis
Publication DetailsType:Proceedings: LREC 2018, Eleventh International Conference on Language Resources and Evaluation, Date: 2018Location: Japan,Refereed Conf./Meeting Proceeding
Conferences/Workshops/Seminars
- DetailsInsight Participantthomas gaillatSponsor OrganisationInfotoday, LocationUnited KingdomStart Date17/10/2017 to End Date18/10/2017