Personalized Adaptive User Interfaces

January 30, 2014
2:50 pm - 4:00 pm
Halligan 102
Speaker: Krzysztof Gajos, Harvard University


I aspire to personalize the Web. Not the content, but the interaction.

In this talk, I will share with you results of four projects, each of which explored a different approach for adapting user interfaces to unique circumstances of the individuals using them. With the Supple project, we demonstrated how to cast interface design as an optimization problem and how to automatically generate user interfaces optimized for the unique motor abilities of individuals with a variety of motor impairments. Users with motor impairments were faster with and strongly preferred the automatically generated user interfaces compared to the default interfaces shipped with today's software. Our SPRWeb tool automatically recolors web sites for individuals with color vision deficiencies (CVDs) such that the resulting color schemes are both differentiable and also evoke similar subjective responses in individuals with CVDs as the original color schemes did in users with typical color vision abilities. In a study exploring the design space of strategies for continuously adapting the structure of a complex user interface to the user's current task, we sought to examine the prevalent assumption that frequent changes to the user interface structure would cause confusion and would be rejected by users. While a number of adaptive approaches failed, our results demonstrated that the Split Interface adaptive approach resulted in improved performance and was strongly preferred by the users compared to equivalent non- adaptive interfaces. Finally, our work on the Adaptive Click-and- Cross interaction technique, was spurred by the observation that the word "adaptive" is used to describe a multitude of different approaches in the context of interactive systems. We demonstrated that these approaches can be synergistically combined: simultaneously adapting to both a user's motor abilities and task improved performance of users with motor impairments compared to adapting to either factor alone.

I will also share results of two of our recent projects, which demonstrate the existence of large cross-cultural differences in aesthetic perception, attitudes toward time and in social decision making. The results of these studies point to additional research opportunities for personalized adaptive design.


Krzysztof Gajos is an associate professor of Computer Science at the Harvard School of Engineering and Applied Sciences. Krzysztof is broadly interested in interactive intelligent systems, a research area that bridges artificial intelligence, machine learning and human- computer interaction. Recent projects pursued by his group touched upon areas such as personalized adaptive user interfaces, computer accessibility, tools that help people learn, creativity support tools, crowdsourcing, and tools and methods for engaging broader publics in research.

Prior to arriving at Harvard, Krzysztof was a postdoctoral researcher at Microsoft Research. He received his PhD from University of Washington and his M.Eng. and B.Sc. degrees from MIT. Krzysztof is a coeditor-in-chief of the ACM Transactions on Interactive Intelligent Systems. He is a recipient of a Sloan Research Fellowship.