ARVUTILINGVISTIKA JA KEELETEHNOLOOGIA TARTU ÜLIKOOLIS
Development of computational linguistics and language technology at the University of Tartu

Summary/Abstract: Development of computational linguistics and language technology at the University of Tartu The article gives an overview of the current research in computational linguistics and language technology at the University of Tartu: research subjects, achievements, and problems. Only research on written language is described, and language resources are mentioned only in brief. In computational morphology, the tools for morphological analysis and synthesis have been robust enough to be used in various practical tasks during the last decade. At present, the subject of most active research is disambiguation. In syntax, a shallow parser has been implemented, and current research is focused on dependency parsing. In semantics, a great deal of effort has been (and is still being) devoted to WordNet and a related task – word sense disambiguation. Recently, tools for named entity classification have been built. Studies in clause semantics focus on spatial movement situations, modelled in frame semantics. In pragmatics, the focus of interest is on modelling dialogues, and in particular, on recognizing dialogue acts (e.g. requests, greetings) as the building blocks of dialogues. In machine translation, the aim is to find language-specific ways to improve statistical machine translation.

• Details
• Contents