Artificial Intelligence Laboratory

Founded in 1992, Artificial Intelligence Laboratory (AiLab) at IMCS, University of Latvia conducts research on natural language processing (NLP) and machine learning (ML) in several directions:

  • Speech-to-text recognition and text-to-speech synthesis for Latvian.
  • Syntactic and semantic parsing for cross-lingual information extraction, question answering, human-computer interaction.
  • Multilingual text generation from abstract meaning representations and from data.
  • Controlled natural language for knowledge representation.
  • Creation of annotated language resources for NLP and ML: machine-readable lexicons, text and speech corpora.
  • Application of deep learning and reinforcement learning in roboticscomputer vision and grounded language learning.

AiLab has recently implemented multiple research and market-oriented projects on large-scale information extraction and speech recognition. In cooperation with National Information Agency LETA, we are handling big data problems arising in the media monitoring business. This successful collaboration allowed LETA to join the EU Horizon 2020 project SUMMA, together with leading European universities, BBC, Deutsche Welle and others. Researchers from AiLab were also involved.

We have developed a text processing system for National Library of Latvia, dealing with part-of-speech tagging and named-entity recognition in a billion word corpus of OCR-scanned texts. We also collaborate with Riga Stradins University, developing platforms for monitoring user-generated content in news portal communities and for analysing transcripts of parliament debates. Likewise, we collaborate with an international defence organisation to develop tools for tracking and countering online disinformation. In collaboration with Riga East University Hospital, we develop Latvian speech recognition and synthesis technology for medical applications.

AiLab researchers collaborate also with foreign researchers from University of Edinburgh, University of Gothenburg, University of Tromsø, Charles University in Prague, Uppsala University, International Computer Science Institute, and other scientific institutions.

Our researchers successfully participate in international NLP and ML competitions like SemEval and TAC KBP. At SemEval-2016, our team achieved the top result in the task on Meaning Representation Parsing, while at SemEval-2017 - the top result in the subtask on AMR-to-English Generation.

Currently we are implementing several applied research projects funded by European Regional Development Fund, Latvian Council of Science, and other funding bodies.

Historical homepage: