UMD Lab Pioneers Methods for Better User Interfaces
With groundbreaking work, the University of Maryland Human-Computer Interaction Lab works to bring information seekers closer to their goals
By Karen Heyman
For the DGRC

UMD's Human-Computer Interaction Lab

• University of Maryland Human-Computer Interaction Lab • Tree Maps • PhotoMesa • "Improving Accessibility and Usability of Geo-referenced Statistical Data" (dg.o2003) • "Data Exploration with Paired Hierarchical Visualizations: Initial Designs of PairTrees" (dg.o2003) • "Helping Users Get Started with Visual Interfaces: Multi-Layered Interfaces, Integrated Initial Guidance and Video Demonstrations" (dg.o2003)

Good interface design, it is often said, should not be noticed.

Users don't want to think about page layout and levels of links, they simply want to find information. Bad interface design is not merely noticed; it is yelled and cursed at by users frustrated by how much time they have lost seeking what should have been obvious. Sometimes the bad design is simply the result of inexperienced or even untalented creators, but often it comes out of incorrect assumptions, from not having asked the right questions before building.

What is the most likely path people will take through a Web site? How many levels down should information be stored? What if the user is visually impaired? Do people from different cultures approach the same visual scheme differently? Will people understand intuitively that they are supposed to press a touch screenÑor do you need text that alerts them that smudging the screen is not only okay, it's necessary?

Researchers at the Human-Computer Interaction Library at the University of Maryland in College Park, Maryland have spent two decades trying to answer those and similar questions. But even more important than answers is the design of the questions themselves. The HCIL's expertise lies in uncovering the thought processes people go through as they seek information.

Says Associate Research Scientist and Associate Director of the Lab, Catherine Plaisant, "We're known for trying to evaluate the interfaces we design through usability studies, controlled experiments. How does the interface change things? Sometimes it only adds complexity and you don't get benefits. Sometimes it's a matter of gaining a few seconds on a task, which saves millions of dollars, such as for telephone operators looking in a directory. Sometimes, it's an issue of 'Can the user actually find it?'"

The ideas the lab has developed have become fundamental to much of human computer interaction.

In the early 80's, lab founder Ben Shneiderman and his colleagues began to work on implementing a concept that had been suggested by Vannevar Bush in his now famous 1945 Atlantic Monthly essay "As We May Think," which first laid out the ideas that would eventually be realized in the Internet. Shneiderman called the method for clicking on text to reach other documents, "embedded menus" or "illuminated links."

In his 1989 conceptual paper laying out the idea of the World Wide Web, Tim Berners-Lee would cite work by the UMd team and several other research groups in proposing a public web of information linked by hypertext.

The University of Maryland lab's work has led to several commercial products and government commissions. In 1988, a public access kiosk designed by several lab members for an exhibit at the Smithsonian Institution featured the first touchable image map for PC-based computers.

One of the lab's most popular ideas is "TreeMap." It has spawned both commercial and open source products. A tree-map is a visual representation of statistical information that automatically re-sizes to show proportion.

Plaisant gives an example, "One map showed, 'The 42 Main Causes of Death'. Each box represents a cause of death, and the size is proportional to the number of people, and color indicates if it's [increasing or decreasing]. You can see at a glance that cancer is down, and Alzheimer's is way up." Less grimly, the tree-map algorithms power an image browser called "PhotoMesa," where photos are hierarchically arranged, and users can zoom in and out of hundreds of tiled images to easily find one picture.

Most recently the lab has received a great deal of favorable press for the International Children's Digital Library, the brainchild of current lab director Ben Bederson and his education professor Allison Druin, his wife. Funded by the NSF and others, the Library is designed to be accessed by children ages 3 to 13. Druin discovered that children ask different questions when they search for books than adults do. They actually seek out books based on how they might make them feel, or search for books by the color of their covers.

Latest DG News

• dg.o 2006 Convenes May 21-24, 2006   • dg.o 2006 Early Registration Ends April 10th! • dg.o 2006 Issues CFP - Tutorials • dg.o 2006 Issues CFP - Workshops • dg.o 2006 features Workshops on:    eRulemaking    GeoInformatics • dg.o 2006 features Tutorial on:    •Social Network Analysis • New DG Team Pursues eRulemaking • IEEE ISI2006 Convenes May 22-24, 2006 • eChallenges e-2006 Issues CFP • DG Research Helps Predict Urban Growth • Swapping Secrets of the Double Helix • UK and DO-Wire Launch e-Gov Best Practices wiki • DG Team Develops "Virtual Agora" for e-Gov • Mapping for Times of Crisis • Exploring Detection of Crisis Hotspots • Report: Mass eMail Campaigns Harmful • Scenario-Based Designs for Stat Studies • US, EU Explore Info Integration • DG Team Develops Digital Interpreter • DG Study Gives Teeth to FBI • Research Smooths Road for Small Businesses • DG Researchers Parsing in Tongues • e-Gov Journal Issus Call for Articles • See all news stories Contribute to dgOnline

Users who are more visually than text-oriented fit in well with the lab's approach. Says Plaisant, "Over the years, we've been trying to develop more visual interfaces, typically statistical information is presented in tables. We've been looking at interfaces, like dynamic maps, that allow people to interact with the data, to specify their query dynamically." For example, if a person were to ask which states had the most retirees, the colors on the map would change in response to the question, immediately highlighting which states had the densest populations of retirees.

One of the most recent NSF-funded projects is GovStat. The lab is working on it in collaboration with the University of North Carolina Interaction Design Lab. (See accompanying article). Plaisant describes some of the interface challenges, "People know good interface design is important, but they don't always know how hard it is to do it well. How do you get everybody to learn to use your interface, consistently. Each site might be well done, but people get lost going from site to site."

As a result, she says, "One part of the project is to help people get started. [Since] people don't use tutorials, how - in as few seconds as possible - do we help people get started?" One solution has been animation, with items such as animated maps.

But visual techniques run into the universal access problem Ñ people who are blind cannot use the maps at all. Although there are alternatives such as screen readers and tactile printers, they are not perfect solutions, explains Plaisant. "Screen readers do well with text, but badly with tables. When tables are in PDF, they get the tables completely wrong Ñ Maryland is the 30th state, and you have to remember to listen to the 30th number."

The lab is now exploring "sonification". "There are techniques to help reproduce 3D sounds through the same headsets the blind use for the screen reader," explains Plaisant.

In this scenario, when a user mouses over a map, a noise gets louder or softer to indicate values. Plaisant says that at a conference on the subject research was presented that says people have a natural bias about soundsÑa high tone is taken to indicate something higher up, a low tone, something lower down. "There are new techniques for sound - up and down, and right and left." However, she cautions, "It may not work, there may be better ways that don't use spatial sound."

Reviewing the lab's twenty years of collaborations with government agencies Plaisant says, "We learn the problems of the government partners, and then go back and say, 'we also work on this, this might help you too.' It brings us back to reality Ñ it's motivating for the students who say, 'Wow, if I do a good job, maybe millions of people will use this.'"