Welcome to the website of the Computational Linguistics group at the Institute for Language Sciences (ILS)!

The Computational Linguistics group focuses on computational accounts of language systems from the point of view of cognitive modelling. Its primary goal is to study connections between state-of-the-art NLP models and linguistic theories of natural language. In what ways do current NLP systems mimic the human mind? To what extent do they understand language similarly to humans? Is the acquisition of linguistic knowledge similar in preschoolers and state-of-the-art talking machines? In what ways can linguistic theory help to improve NLP models?

Together with developing practical NLP applications, research in the group pays special attention to formal models, to experimental results with language speakers, to facts of language acquisition, and to “long-tail” phenomena that are often challenging for state-of-the-art NLP.

The researchers in the group focus on computational cognitive models of sound (Adriaans, Nazarov), structure (Deoskar, Fowlie) and meaning (Dotlacil, Nouwen, Roberts), with a special interest on AI and Reasoning (Abzianidze, Paperno, Winter).

Research in the team operates within the Institute of Language Sciences (ILS). As part of the Linguistics section of the Department of Languages, Literature and Communication (TLC), the group members teach in the Linguistics bachelor/master and AI bachelor/master programs.