Although Xiaoli Fern graduated from high school two years earlier than is typical in China and started college at the age of 16, she is quick to explain that she was not a prodigy.
“I’m not one of those people,” she laughs. “I took every year of school. But both my parents worked, so it was easier to just start me in school early.”
Nevertheless, she didn’t have any trouble academically but recalls it was a bit harder to find friends being the littlest kid in her grade. Her father was in the military, so her family moved frequently making childhood a lesson in adapting.
“Humans are pretty flexible,” she says. “You get used to something and then it will take a little bit of bending to get used to something else.”
So, when it came to leaving her family to come to the US for graduate school at Purdue University, she was undaunted. However, the difference between bustling Shanghai and the small town of Purdue, Indiana was still a “pretty big shock” she says, initially viewing Purdue as a corn patch in the middle of nowhere.
“I’m used to it now. I can’t handle big city life anymore,” she says.
Adapting and learning is what she expects from computers as well. A computer scientist who studies artificial intelligence, her main area of research is active machine learning — designing systems in which the computer program seeks information from the user.
Much of her research focuses on a powerful machine learning tool that automatically organizes large amounts of data into what are called clusters. Although it is often used by scientists for applications such as identifying genes that perform similar functions, she says its broad use is hampered because there is no mechanism for novice users to guide the clustering.
“It’s a very useful technique that can be used universally, and what I’m doing is trying to make this approach more accessible for people by developing a machine learning system in which the software asks questions of the user to help define the categories,” she says.
One application will help ecologists classify bird songs from recordings taken in the wild with the goal of understanding the meaning of different song types. Another application of her work will result in more intuitive user interfaces like those found in smartphones.
Her outstanding research and teaching has earned her a prestigious CAREER award from the National Science Foundation, given to top researchers at the beginning of their career.
But Fern credits her collaborations with her success.
“Collaborations keep the research alive. You don’t have to imagine what your work can be used for, you have real applications that push you in directions that you wouldn’t have thought of on your own,” she says.
Among her collaborators she counts all the students that work in her lab — graduate students, undergraduates and high-school students.
“To have students at each level is exciting because they can learn from each other, and the younger ones can see that it is really not that hard — that college and graduate school is something within their reach,” she says.
Fern also enjoys hosting undergraduates on summer research projects through the Computer Research Association’s Committee on the Status of Women in Computing Research (CRA-W), who have worked on the bird song project. It’s a partnership, she says, that seems to benefit everyone.
“The applied nature of the work and the connection with the real environments is very important to a lot of female students,” she says.
She and husband, Alan Fern (also computer science faculty), are also very busy at home raising three young children whom she is trying to teach Chinese — and she is hopeful that the family will get a chance to have a sabbatical to China where her father still lives.
—By Rachel Robertson