Making High-Tech Eye Candy
By Jennifer Fisher Wilson
This is the reality today in the field of computer science: compared to men, fewer women earn bachelor's, master's or doctoral degrees in computer science, and the numbers tend to decline from undergraduate to graduate levels. According to the Digest of Education Statistics compiled by the U.S. Department of Education, public high-school girls and boys have earned an equivalent number of computer-science course credits from 1982 through 2000. But parity becomes inequality after high school. In 1984 the percentage of all computer-science bachelor's degrees earned by women peaked at 37 percent, and has since declined to 28 percent by 2002. In 1984 women accounted for 29 percent of computer-science master's degrees; although the trend has improved slightly to 33 percent in 2002, women still earn fewer master's degrees in the field than men. And in 1984 women accounted for just 10 percent of doctoral degrees in computer science, and 23 percent in 2002 — encouraging growth, but still far fewer than men.
Dianna Xu, a new assistant professor in the computer science program at Bryn Mawr, experienced firsthand the narrowing pipeline of women in her field. She was the only woman who started computer-science graduate studies in 1996 at the University of Pennsylvania and finished with a doctoral degree. "There's a lot of speculation about why so few women are in the field, including prejudices, social pressures and a lack of role models. As it stands, girls simply don't think of computer science as a possibility. Most end up in the field by accident," Xu says.
Xu herself never planned to major in computer science. She took a computer course during her first year at Smith College only because a last-minute opening in her schedule needed to be filled. "I figured computers are useful, maybe I'll just take a course. But once I took that computer programming course, I never looked back," Xu recalls.
Field of Choice
At Bryn Mawr, Xu wants to make computer science a course of study students choose to pursue, instead of discovering it by accident. She hopes to serve as a role model and encourage more female students to see the thrills of working in computer science. "With computers, there's an excitement that comes from seeing your work happen onscreen. The process may be tedious and time consuming, but in the end, you see the 'eye candy.' Then it's worth all of the trouble," she says.
Certainly, she admits, interest in computer science is not universal. She sees that in her students. "Some can hardly stand computer programming because it can be such precise, tedious work. But others get an overwhelming sense of accomplishment when they solve a programming problem. Those are the ones who stay in the lab until three in the morning to get it right. Those are the ones who could become computer scientists," she says.
Xu coaches some of these students in interscholastic computer competitions focused on a team approach to solving complex programming problems. Bryn Mawr is one of just two women's colleges with programming teams in the Association for Computing Machinery International Collegiate Programming Contest, the largest and most prestigious programming competition in the world, Xu explains. Among 161 teams, the Bryn Mawr team recently placed in the top third.
Graphics and Vision
Xu's academic research provides compelling examples of the important roles computers can play in modern life. Her efforts are focused on computer graphics — creating images on the computer from scratch — and a related discipline called computer vision — creating images from existing pictures.
She is perfecting a computer graphics tool called triangular spline surfaces. These surfaces are used to solve geometric problems that arise in architecture, manufacturing, graphics modeling and animation, among other fields. For instance, Xu says, this design approach is used to model new cars on the computer. Computer scientists create a complete model for how a car will look and how it will function, including such characteristics as streamlining, stability and fuel efficiency. By running the model through computer-simulated tests, the scientists can adjust the model to achieve the most desired look and function. Once the designers have created an accurate digital model, the car manufacturer will build a prototype and put it through real-life testing.
"Even then, once you have an accurate digital model, computers can be used to control the actual manufacturing phase and reduce manufacturing errors. For instance, the computer can precisely guide an electronic arm so that a piece of metal will have just the right curvature," Xu says. In particular, she is developing a new algorithm to design smooth, modifiable triangular spline surfaces suitable for designing complex, three-dimensional shapes.
She is also involved with the University of Pennsylvania Museum of Archaeology and Anthropology in a computer-vision project to reconstruct the ruins of the pre-Columbian temple site of Tiwanaku in Bolivia. Xu and other computer scientists are creating virtual three-dimensional walkthroughs of the ceremonial center using inexpensive digital cameras operated by archaeologists in the field.
"The archaeologists think the ruins are pieces of a temple, but they don't know how the stones fit together. We suggested they could construct the whole site using digital models based on stereophotographs. If the computer models are accurate, people could potentially recreate the temple, using the computer pictures of the rocks to solve the jigsaw puzzle." Xu says.
Looking forward, Xu says she is excited about how much potential the field of computer science holds for changing the way people experience the real and imagined world. Computer graphics, for instance, may soon be able to achieve virtual reality. "It could be really awesome. Virtual reality has applications in practically everything — virtual conferencing, military training, medical education, the movie industry and, in particular, gaming," she says. Xu hopes that in the future, more women will be a part of making similarly exciting contributions to the field of computer science.
Jennifer Fisher Wilson is the science writer for Annals of Internal Medicine.