Choice Architecture for Human-Computer Interaction
How can a better understanding of human choice and decision making benefit design of and research on interactive systems?
Users of computing technology are constantly making small choices and larger decisions. Research areas in a variety of disciplines have together produced a vast collection of knowledge about the different ways in which people make choices and about ways of supporting them. But so far, only isolated parts of this knowledge have been used to improve people’s interaction with computing technology.
In collaboration with several colleagues, I have been working since 2008 on ways of synthesizing and complementing this research to make it useful for people who design and study interactive computing systems. This work has resulted in the publication of the book-length journal article Choice Architecture for Human-Computer Interaction, which is intended to serve as a foundational resource for everyone who is interested in using computing technology to help people make everyday choices.
This publication benefitted from experience with the preparation and presentation of several tutorials in this area at CHI and IUI conferences (listed on the page tutorials) and of a chapter for the 3rd edition of the Human-Computer Interaction Handbook.
Recent talks on this topic are listed on the page with Invited Talks.
How can interactive systems help choosers explore in several directions in parallel?
A typical limitation of technology for supporting people’s everyday choice processes is that they make it unnecessarily difficult to explore options in two or more directions in parallel. In the projects GLOCAL, EventMAP, and Apps for Your Car, my group developed the general interaction paradigm that we first called parallel faceted browsing and have now generalized to parallel exploration.
This a web-based app using this paradigm is serving as a key component of the platform and showcase application of the multinational flagship project 3cixty of EIT ICT Labs. Several publications about earlier versions will be found on the publications page.
Binocular View of Interactive Intelligent Systems
How can we achieve a better synergy among different approaches to research on interactive intelligent systems?
Technology involving various forms of machine intelligence already serves useful functions in a wide variety of interactive computing systems, and the open-ended future possibilities are being explored in a number of research communities. These communities can learn a lot from each other about general issues that arise when human and machine intelligence interact.
The special issue of AI Magazine titled “Usable AI”, focused on questions of usability. As a broader effort, the late John Riedl and I founded the ACM Transactions on Interactive Intelligent Systems, which publishes relevant research from a wide range of areas.
Earlier Research Topics
Publications on the following research topics will be found on the publications page:
- User-adaptive systems in general
- User-centered design of systems based on semantic or language technologies
- Non-mainstream approaches to recommendation
- Ubiquitous and mobile computing
- Adaptation of human-computer dialogs to resource limitations
- Application of decision-theoretic inference, decision making, and learning techniques, where applicable, to the above problems
- Pragmatics of natural language dialog, ascription of knowledge and belief