Mark Fišel to deliver inaugural lecture on natural language processing
On Wednesday, 22 January at 16:15, Mark Fišel, Professor of Natural Language Processing at the University of Tartu’s Institute of Computer Science, will be delivering a lecture in the university assembly hall on natural language processing solutions that do not require annotated examples or linguistic knowledge encoding. Fišel, who compares his field of work to the search for a free lunch, has called his English-language lecture just that: ‘Free Lunch’.
According to Mare Koit, a language technology specialist at the University of Tartu, natural language processing has made significant progress thanks to the adoption of machine learning and statistical methods and the availability of large language corpora. “Mark became active in the field as a student and has since produced significant results in the field of machine translation in particular,” she said. “Under his leadership, the Chair of Natural Language Processing is conducting a number of research projects and working with business partners. For instance, Estonian was added to the Common Voice project by Mozilla, which is used to collect freely usable speech datasets. Mark’s colleagues and students value his ambitious ideas, how hard-working he is and how much energy and enthusiasm he has.”
University of Tartu language technology specialists study and develop the basic technology of natural language processing, which is largely independent of language. At the same time they develop applications and natural language processing components for Estonian end users, offering technological support for the Estonian language.
Fišel sees natural language processing as a complicated and exciting branch of artificial intelligence – mainly because nobody knows exactly how language functions and current solutions for individual tasks, such as machine translation and text analysis, are imprecise and incomplete. “The majority of existing language technology solutions are based on large amounts of language resources,” he explained. “As a result, the situation becomes more complicated in the case of languages with fewer speakers and fewer resources, such as Estonian.” His lecture will address such topics as machine translation, translation quality estimation, style adaptation and grammatical error correction.
Fišel obtained a PhD in Computer Science from the University of Tartu in 2011 and conducted postdoctoral research at the University of Zurich from 2011–2015. He has been the head of the Chair of Natural Language Processing at the University of Tartu since 2015.
Mark Fišel, Professor of Natural Language Processing, Head of Chair and Vice-Director of UT Institute of Computer Science, +372 737 5430, mark.fisel [ät] ut.ee
Mare Koit, UT Language Technology Specialist, +372 737 5464, mare.koit [ät] ut.ee