REDMOND, Wash., March 27, 2003 — Desney Tan, a graduate student at Carnegie Mellon University in Pittsburgh, and Tomer Moscovich, a graduate student at Brown University in Providence, R.I., are getting an education in tackling real-world problems. Although he’s still at least a year away from earning his doctorate in computer science, Tan has already helped uncover why male and female computer users navigate 3-D environments differently. Moscovich, meanwhile, has helped establish the viability of freeform ink annotations for digital documents, something that could well transform the way people interact with documents in the future.
Those were the key findings of two original research projects that Tan and Moscovich collaborated on as interns at Microsoft Research. The findings are compelling enough that papers on both topics were accepted for publication at the upcoming CHI 2003 Conference on Human Factors in Computing Systems. Slated for April 5-10 in Fort Lauderdale, Fla., CHI 2003 is a leading international forum for the exchange of ideas and information about computer-human interaction (CHI).
CHI is an important professional conference for many researchers. Now in its 21st year, it brings together designers, computer scientists, psychologists and anthropologists, making it a key event for people working in the field of human-computer interaction. Besides listening to presentations of significant papers and seeing the latest technology demonstrations, attendees have an opportunity to network with colleagues and heighten the public’s awareness of good design.
The competition is stiff for researchers hoping to have their work reviewed and recognized by their peers at this venue. Of the 470 papers submitted to CHI 2003, only 75 were accepted for publication. The fact that interns contributed to six of the nine full-length technical papers accepted from Microsoft Research demonstrates that students like Tan and Moscovich play an increasingly vital role in the process.
“Interns are critical to the work we do,”
says Mary Czerwinski, Tan’s mentor and the senior researcher and manager of the Large Display User Experience group at Microsoft Research.
“Interns really complement our researchers. And with someone like Desney, who is so technically adept, it’s very much a collegial relationship. I learn as much from him as he learns from me.”
Interns are indispensable, agrees David Bargeron, a researcher in the Document Processing and Understanding group at Microsoft Research and the co-author of another paper accepted at CHI 2003.
“These are students in the vanguard of people developing their research skills and talents, and as interns, they are finding their own style and direction,”
“But their presence is also important to us at Microsoft Research. On the creative side, we get fresh ideas and approaches that energize our work. Staying connected with academia also gives us a chance to see who’s doing cool work and who has interesting insights.”
Key Role of Interns in Research
This year’s CHI conference will find Tan at the podium, delivering two papers that he co-wrote with his Microsoft Research colleagues. The first, titled
“Women Go With the (Optical) Flow,”
stems from collaborative research he did with Czerwinski, a psychologist, and George Robertson, a computer scientist at Microsoft Research.
Their study discovered the reasons behind gender differences in how users navigate a virtual 3-D world. According to Czerwinski, the novel findings mark a significant research contribution to both the discipline of human-computer interaction and the field of psychology.
The optical-flow project dates back to Tan’s first internship at Microsoft Research in summer 2000. He recalls being surprised when, instead of just assigning him to a project, researchers asked what
he’d like to do and what resources he would need. The approach contrasted sharply with his other intern experiences.
The second big difference Tan noticed was the open-door policy.
“Microsoft Research has some of top researchers in the world, and at any given time I can knock on someone’s door and say, ‘I’m working on this problem, here’s what I came up with, what do you think?'”
“Researchers are more than happy to sit down and brainstorm or sketch ideas out on a white board. Probably more impressive is how those same researchers will come to the interns’ offices and ask what we think about a problem they’re working on. I’m impressed by the open environment and the atmosphere of collaboration.”
Tan also collaborated with Czerwinski on a second paper he will deliver at CHI 2003. Titled
“Information Voyeurism: Social Impact of Physically Large Displays on Information Privacy,”
it’s one of five short papers from Microsoft Research accepted for publication. The study in that paper proved the hypothesis that people are more inclined to consider private information public if it is presented on a large display. The likelihood that others will read the content of, for example, an e-mail message, has implications for compromised privacy. By quantifying this effect, the researchers lay groundwork for designing software with user interface conventions and other mechanisms that better protect personal data from prying eyes.
Both studies dovetail with Tan’s diverse research interests, which range from human cognition and memory to large displays and 3-D interaction. The internships themselves have opened new doors for him, such as publishing at CHI. The intern experience has also defined Tan’s research direction and strengthened his resolve to continue in research once his Ph.D. is in hand.
“Essentially, my goal is trying to find the next big step in human-computer interaction,”
says Tan, 27.
“I want to focus my research on the need to change our ways of thinking about the interface.”
Hands-On Work Makes Research Experience Real
Students who land internships at Microsoft Research typically spend three months at the company’s Redmond campus, working 40-hour weeks alongside the staff researchers. Typically, the arc of a summer internship covers a few weeks of brainstorming and whittling down ideas, followed by multiple iterations of code to implement those ideas, which then leads to revisions in the design and implementation. Often, the final month is devoted to testing the research by putting the results in front of actual users.
That’s essentially what brought Tomer Moscovich back for a second internship at Microsoft Research last summer. Moscovich, who’s working on his doctorate in computer science at Brown, says that what makes his work interesting is not so much the algorithms but the human element. Brown’s computer science department, however, doesn’t have the same focus on studying user behavior that Moscovich experiences working within a software company like Microsoft. Thus, he was eager to return to Microsoft Research and learn how to test the reactions of users to new ways of interacting with computers.
“Aside from picking up more technical skills, like using C#, the main thing I learned was how to design and run a user study, and I learned a lot about data analysis,”
says Moscovich, 25.
“Knowing how to test systems and user reactions will come in very handy in the future if I intend to do more of this work.”
Bargeron, who serves as Moscovich’s mentor at Microsoft Research, notes that designing and organizing user studies represents a great deal of logistical work. But he agrees that it’s an important aspect of what a computer science researcher does. Another key experience that Bargeron likes to foster in interns is presenting papers on their work, which Moscovich will do at CHI 2003.
Moscovich and Bargeron collaborated on a research project that focused on preserving the intended meaning of annotations (such as underlining and margin notes) to digital documents as the layout of a document changes. Beyond just proving that it’s feasible, the two reached some encouraging conclusions as far as user expectations, which should help designers and system builders support freeform annotation more effectively in the future. Their findings are reported in a joint paper titled
“Reflowing Digital Ink Annotations.”
As with most Microsoft Research projects that involve interns, Moscovich was encouraged to take part in all aspects of the process. The first part of the project involved brainstorming ideas for reflowing annotations, developing an implementation system and creating an annotation tool. The second part focused on planning the user study, which included deciding what to measure and how to measure it, and designing a questionnaire to gauge users’ level of understanding. The final weeks were spent analyzing the results of the user study and drafting the paper that Moscovich will deliver at the conference.
“Being an intern let me experience what it’s like to be a researcher,”
“In some ways, it’s similar to the research you do in academia, but the focus is quite different. Most of my work previously stopped at proof of concept, whereas at Microsoft Research the ultimate goal is to integrate research into real products.”
For his part, Bargeron highly recommends the intern experience to his colleagues at Microsoft Research, pointing out that it’s an opportunity to work with some of the best up-and-coming minds in academia.
“In a sense, you could regard interns as extra labor, but at this level, where it wouldn’t be sufficient to hire a contractor or a software engineer, you need someone who understands the work and can engage with it creatively and take it forward,”
“It’s a chance for me to take some ideas that have been waiting for time and resources and make them happen. And then you find someone like Tomer who really gets the idea and can run with it, and it’s an ‘all-boats-rise’ situation. In other words, we all win.”