Adaptive Interaction Techniques for Sharing and Reusing Design Resources

Brad Myers   Brian Lee , Stanford HCI

Seminar on People, Computers, and Design
Stanford University December 7, 2007

Today's designers generate content both on paper and online. Designers spread their work over physical and digital media, each of which has powerful – but distinct – sets of affordances. Recent work suggests that augmented paper interfaces can marry the ubiquity of paper interactions with the ease of search, annotation, and presentation afforded by digital representation. This dissertation examines novel ways to support and augment the practice of design through sharing and reappropriation of digitally captured design content.

The thesis of this dissertation is that an ecology for design that integrates augmented paper and digital tools can facilitate collaboration between designers and improve the visibility of design resources. Our contributions are twofold: we study actual use of augmented paper tools for capture and access of design content, and we design selection and presentation algorithms for proactive display of design materials.

To investigate the potential value of augmented tools for design, we developed the iDeas design ecology, which integrates physical notebooks with a digital faceted metadata browser that offers explicit annotation and sharing mechanisms, and conducted four studies with student design teams. Our findings indicate that while there are clear benefits to use, such as increased excerpting and sharing of design material, naïve sharing mechanisms carry significant perceived costs to adoption, including concerns over permissions, which depress usage.

The findings from these studies motivated our second tool, Adaptive Ideas, which explores the use of implicit sharing mechanisms to improve visibility of example design resources. We describe an optimization-based approach to selecting and presenting design material adaptively, using decision-theoretic selection, designer specification, and end-user preference as inputs. Results from two laboratory studies of an example-based web page builder indicate that proactive presentation of examples is useful in helping designers explore and understand spaces of design alternatives.

Brian Lee is a Ph.D. candidate in Computer Science at Stanford University. He studies Human-Computer Interaction, and is advised by Scott Klemmer.

View this talk on line at CS547 on Stanford OnLine or using this video link.

Titles and abstracts for previous years are available by year and by speaker.