Natural Language Processing (NLP) is the task of using computer programs to analyze, understand and generate natural language. Our research concerns both basic algorithms for language AI and the use of such algorithms to solve various practical tasks and problems involving language data.
Natural language processing has become a key competence in modern information technology, with applications in virtually all areas of the private and public sectors that deal with language data. In general, NLP applications are about extracting useful information from language data, and in particular from text material. Of course, what counts as useful information depends on the user's needs, but examples include:
In order to perform these types of NLP tasks in the best possible way, we need algorithms that can learn and understand languages in the same way as humans do. Recent years have seen significant breakthroughs in language AI, which have produced results that in some cases approach human performance. This type of modern language AI has the potential to radically improve the handling of language data in both the private and public sectors. In particular, we are committed to support Swedish actors in developing and integrating the latest NLP technologies into their operations.
We have extensive experience in applied NLP research, and we have leading expertise in finding NLP solutions to business-critical problems. Our customers and partners are in academia as well as in the private and public sectors.
Seminar series for people who work with, or are interested in, NLP.
Language models for Swedish authorities – this project develops and applies large-scale language models for Swedish authorities
Strategic AI-collaboration with the Swedish Tax Agency – this project investigates how NLP solutions can be used to streamline and improve the Swedish Tax Agency's operations.
Linguistic Explorations of Societies – this project develops and applies NLP solutions to study public opinion and society through analysis of open data from the internet.
KnowIT Fast – this project seeks to improve condition monitoring in the process industry by integrating sensor data with language technology
GPT-SWE: A foundational model for Swedish NLP – this project develops and applies large-scale language models for Swedish