Hot and cold - universal or language-specific? (2009-2011, hosted by the Department of Linguistics, Stockholm University) studies the conceptualization of temperature in natural languages as reflected in their systems of central temperature terms.

Distributionally derived grammatical analysis models (2009-2011) investigates the confluence of formal linguistic description and distributional models of linguistic behaviour.

COMPANIONS (2006-2009) COMputational PersonAl Networked InformatiON Sources concentrates on specific groups who would have their lives significantly improved by computational interfaces we call COMPANIONS, connected by messaging to ambient devices and the WWW, and using Human Language Technology (HLT) and intelligent agent modelling technologies now ready for integrated exploitation of this kind.

Attityd i text (models for attitudinal expression in text) (2006-2008) studies attitudinal expressions in human language, specifically text and written discourse. The project aims to identify and analyze attitudinal expressions by their textual characteristics.

KARMA (2006-2007) is a joint research project between Karolinska Institutet's department for endocrinology, Mando AB and the Interaction Laboratory of SICS. The project is funded by Vinnova and targets the development of a character-based therapeutic application for eating disorders.

ENCORE (2003-2006) extended the notion of collaborative filtering by (1) investigating combinations of content-based and social filtering, (2) finding new ways of presenting and explaining recommendations, and (3) using collaborative filtering in decentralized systems.

DUMAS (2001-2004) investigated and developed dynamic and adaptive multilingual interaction techniques that can learn and adapt their functionality to different users, situations and tasks. These challenges were addressed by integrating research on robust text processing, flexible dialogue management, advanced user interface techniques and dynamic user modelling.

Stochastic Pattern Computing, part of the RWC Project (1993-2001), studied the mathematical foundations of humanlike flexible information processing methods that compute with large random patterns (or high-dimensional random vectors), rather than with numbers and pointers. The achievements of the project led to the development of the Random Indexing technique.