University of Gothenburg
Research groups

Research groups

Within CLASP there are four research groups.

  • Language and Perception Research Group
  • Dialogue Research Group
  • Machine Learning, Cognitive Modeling, and Natural Language Processing Group (MLCMNLP)
  • Type Theory Research Group

Language and Perception Research Group

In the Language and Perception research group we are looking at formal and distributional models (and anything in between) of language used by situated agents interacting with each other and with the physical world around them through action and perception.
We investigate areas such representations of meaning in computational approaches to language, action, and perception, for example of spatial descriptions, generations and interpretation of scene description, multi-modal communication, situated dialogue systems, and other.

The group runs a bi-weekly reading-group and offers two standing PhD courses:

  • Language, Action, and Perception
  • Representation of Meaning 

In 2018 we also organized (together with the Dialogue Research Group) Workshop on Dialogue and Perception.

Members of the Group:

Simon Dobnik (group leader)
Robin Cooper
Staffan Larsson
Shalom Lappin
Christine Howes
Ellen Breitholtz

PhD students:
Wafia Adouane
Sylvie Saget
Vladislav Maraev
Bill Noble

and others.

Please find more information on the group's website.

Dialogue Research Group

We study dialogue from a number of different perspectives, including experimental, computational, formal and qualitative methods.

We have a dialogue reading group which meets fortnightly, please see our website for more details.

The main people in the dialogue research group are:

Christine Howes (group leader)
Ellen Breitholtz
Robin Cooper
Staffan Larsson

PhD students:

Alexander Berman
Vlad Maraev
Bill Noble
Sylvie Saget                                                                                              

In addition, the following members of CLASP have research which overlaps with our area:

Simon Dobnik


In the "Machine Learning, Cognitive Modeling, and Natural Language Processing" (MLCMNLP) group, we are looking at bringing cognitive modeling and theoretical linguistics together with corpus-based, machine learning approaches to both traditional and recent natural language processing problems.  We cover a variety of activities with a particular emphasis on language resource development and applications of human-collected data, be it annotations or experimental results from psycholinguistic research.

Members of the group: In the "Machine Learning, Cognitive Modeling, and Natural Language Processing"

  • Asad Sayeed (group leader)
  • More or less everyone else in CLASP (faculty, Ph.D. students, postdocs) who is interested in machine learning approaches and human cognitive plausibility; these days this is literally just everyone.

Ph.D. students (of whom Asad is first supervisor)

  • Vidya Somashekarappa
  • Axel Almquist

Ph.D. courses

  • Machine Learning, Cognitive Modeling, and Natural Language Processing (standing project/reading course)
  • It's evaluation's world, we just live in it (to be offered second half of Fall 2020)

International collaborations

  • Devdatt Dubhashi's group at Chalmers
  • Yuval Marton (Bloomberg/University of Washington)
  • Vera Demberg's group at Saarland University
  • . . . and others, we are always seeking out collaboration opportunities.

Funded projects

  • Gothenburg Research Initiative for Politically Emergent Systems (GRIPES) -- Marianne och Marcus Wallenbergstiftelsen, WASP-HS

Type Theory Research Group

The Type Theory group is devoted to the study of Type Theoretical methods for NLP and Formal Semantics. The group is led by Stergios Chatzikyriakidis.

Group members:

  • Stergios Chatzikyriakidis
  • Jean-Philippe Bernardy
  • Robin Cooper
  • Staffan Larsson
  • Vlad Maraev
  • Bill Noble

The group’s activities involve invited talks by prominent researchers in the field, organizing workshops on Type Theory and exploring connections between Type Theory and Probability and/or Machine Learning.