Team Members: John Stasko, Dave McColgin, Todd Miller, Chris Plaue, Zach Pousman

Movie - Video overview of project (74 megabyte MPEG)

People have a great deal of personal information (financial, weather, news, traffic, communication, etc.) that they'd like to maintain awareness of, but current methods for doing so are too "heavyweight". That is, the methods use too much valuable time, as in interruptions, or space, as in desktop real estate.

We are seeking to create information awareness techniques and systems that are opportunistic, peripheral, and elegant. More specifically, we want information awareness tools to be

Personalized and consolidated - People should be able to choose information that is personally relevant, and the display of this information should be brought together to promote efficiency.
Elegant and peripheral - Displays of the information should blend smoothly into a person's home or work area and enhance a person's environment. They should not detract from other work spaces such as the person's computer desktop.
Senstive to privacy - Displays should abstract or hide details of sensitive personal information such as financial accounts.

Our approach creates tools that are a form of "Information Art." That is, we use abstract scenes or "virtual paintings" in which the images in the scenes change subtly to convey the state of an information resource. An individual creates their own painting, specifying both the data of personal interest and the images to represent that data. Our approach is similar to, but different in important ways, from that of the PLAY Research group's "Informative Art" project. In their system, portions of existing abstract artworks are modified to represent changes in data being monitored.

Our work began with an extensive survey of people's needs and wants in this area. We interviewed many people and asked about their desires to maintain awareness of personal information. We used the data gathered from that study to develop an initial concept prototype system, called the InfoCanvas. This work was reported at the AVI 2002 Conference.

Experiences gathered from that effort helped us focus on the InformationArt creation process. Currently, we are using an approach that provides different painting "themes" such as a beach scene, a rain forest, a city market, a night skyline, and an aquarium (as shown to right). We have created extensive sets of images (objects) for these scenes. Each InfoCanvas scene is described through an XML file that specifies both the information to be monitored, its graphical representation, and the mapping from the information to the graphical object state. The InfoCanvas is a java application that reads this XML specification file, displays the appropriate picture, and updates object representations in the picture as information changes. The InfoCanvas uses data harvesters (agents that scrape web pages) to gather the information being monitored. The system is described in more detail in a paper at the Ubicomp 2004 Conference.

We also conducted a "glanceability" study in which we analyzed people's abilities to glance at an InfoCanvas display and digest the information being communicated. We compared a Canvas to a web portal display and a textual display of that same information, and we found that people recalled statistically significantly more information with the InfoCanvas. This study was described in a paper at the Graphics Interface 2004 Conference.

In a longitudinal study, we deployed the InfoCanvas to a number of trial users to run for a month in their offices in order to understand how they use the system and what they think of it. We conducted extensive interviews with the participants to learn more about their impressions of the system and how the InfoCanvas does or does not assist with information awareness.

Presently, we also are designing an interactive front-end tool that will allow people to design and create their own InfoCanvases.

InfoCanvas Home | Approach | Downloads | Flash Overview | In the News