Workshop: Theory in Computational Interaction

Organisers Antti Oulasvirta, Thomas Langerak (Aalto U), Philipp Wintersberger (IT:U), Patrick Ebel (HPI), Michal Miazga (U Leipzig), Danqing Shi (U Cambridge), Helena Frijns (IT:U), and Andrew Howes (Exeter)

Motivation The primary goal of this workshop is to chart the landscape of theories of computational interaction. We aim to identify commonalities and differences, and to better understand what each theory does best. Based on these discussions, we seek to establish shared goals for the further development of such theories.

Call for participation: We invite speakers to introduce a theory in an 8-minute pitch. While the format is flexible, each talk must address the following four questions:

1. How does the theory represent humans? Explain its core assumptions and assess them from a psychological perspective.
2. How does the theory represent the computer or AI? Explain its core assumptions and assess them from a computer science perspective.
3. What mechanisms link the two (human and computer) and give rise to possibly emergent outcomes in human—computer interaction?
4. Which computational problems in interaction can be rigorously defined using the theory, and to what extent has progress been made toward solving them?

Outcomes The workshop intends to release a white paper outlining a roadmap toward better theories in this space, with the following elements:

1. requirements for computational theories (what should they do)
2. roadmapping from where we are now toward that goal.

The white paper will be published and later developed further in a Dagstuhl seminar in October dedicated to computational rationality. The workshop will also study the possibility of making a special issue proposal for TOCHI with the working title “Special Issue on Theories of Human-AI interaction”.