Bayesian Methods for Interaction and Design

Introduces Bayesian methods and their implementation in application ranging from pointing-based interfaces to modelling cognitive processes.

John H. Williamson (Edited by), Antti Oulasvirta (Edited by), Per Ola Kristensson (Edited by), Nikola Banovic (Edited by)

9781108834995, Cambridge University Press

Hardback, published 25 August 2022

400 pages
23.5 x 15.7 x 2.6 cm, 0.67 kg

'A succinct and very readable account of network flow algorithms covering the classics and the latest developments. The perfect book for a course on network flow algorithms and a reference for the state of the art. It will be a frequently used addition to my bookshelf.' Kurt Mehlhorn, Max-Planck Institute for Informatics

Intended for researchers and practitioners in interaction design, this book shows how Bayesian models can be brought to bear on problems of interface design and user modelling. It introduces and motivates Bayesian modelling and illustrates how powerful these ideas can be in thinking about human-computer interaction, especially in representing and manipulating uncertainty. Bayesian methods are increasingly practical as computational tools to implement them become more widely available, and offer a principled foundation to reason about interaction design. The book opens with a self-contained tutorial on Bayesian concepts and their practical implementation, tailored for the background and needs of interaction designers. The contributed chapters cover the use of Bayesian probabilistic modelling in a diverse set of applications, including improving pointing-based interfaces; efficient text entry using modern language models; advanced interface design using cutting-edge techniques in Bayesian optimisation; and Bayesian approaches to modelling the cognitive processes of users.

Preface Nikola Banovic, Per Ola Kristensson, Antti Oulasvirta and John H. Williamson
Part I. Introduction to Bayesian Methods: 1. An introduction to Bayesian methods for interaction design John H. Williamson
2. Bayesian statistics Alan Dix
Part II. Probabilistic Interfaces and Inference of Intent: 3. Bayesian information gain to design interaction Wanyu Liu, Olivier Rioul and Michel Beaudouin-Lafon
4. Bayesian command selection Suwen Zhu, Xiangmin Fan, Feng Tian and Xiaojun Bi
5. Probabilistic UI representation and reasoning in touch interfaces Daniel Buschek
6. Statistical keyboard decoding Dylan Gaines, John Dudley, Per Ola Kristensson and Keith Vertanen
7. Human–Computer interaction design and inverse problems Roderick Murray-Smith, John H. Williamson and Francesco Tonolini
Part III. Bayesian Optimisation in Interaction Design: 8. Preferential Bayesian optimisation for visual design Yuki Koyama, Toby Chong and Takeo Igarashi
9. Bayesian optimisation of interface features John Dudley and Per Ola Kristensson
Part IV. Bayesian Cognitive Modelling: 10. Cue integration in input performance Byungjoo Lee
11. Bayesian parameter inference for cognitive simulators Jussi P.P. Jokinen, Ulpu Remes, Tuomo Kujala and Jukka Corander
Part V. Appendix. Mathematical background and notation John H. Williamson.

Subject Areas: Ethical & social aspects of IT [UBJ]