» Yizhou Sun brings interdisciplinary approach to data mining

Yizhou Sun brings interdisciplinary approach to data mining

By Shana Vu

Mining Madness

Yizhou Sun’s eyes light up when the topic of data mining comes up.  

One of only four female computer science professors at UCLA, Sun may be new to the department, but data mining has been the focus of her career since her undergraduate days. “With data mining, you can solve real world problems no matter what the domain,” she said.

Data mining, the process of collecting and analyzing data to find meaningful patterns and to make predictions, is described by Sun as a universal tool that has applications to a variety of different fields, from healthcare to social science.

“Data doesn’t necessarily mean knowledge, so by using computer science to power statistical analysis, we can build models that help us understand the world and predict the future,” she said.

From Beijing to Boston

Sun obtained her undergraduate and master's degree at Peking University, a major research institute located in Beijing, China, where she double majored in computer science and statistics.

Excited at the prospect of conducting research in computer science in the United States, Sun entered a Ph.D. program at the University of Illinois at Urbana Champaign (UIUC), obtaining her doctorate in 2012. It was at UIUC that Sun met her advisor who was a pioneer in the emerging field of data mining.

Before coming to UCLA in July of this year, Sun spent three and half years as an assistant professor at the College of Computer and Information Science of Northeastern in Boston. During her time there, she won an NSF career award for her work on mining heterogeneous information networks with human factors.

An Unconventional Approach

Reflecting on her time at Northeastern, Sun said she particularly enjoyed the school’s decidedly non-traditional approach to computer science. “Computer science is itself a college, which is very interdisciplinary, , you had computer scientists, political scientists, and physicists crossing paths and working together all the time,” she explained.

Northeastern encouraged diversity in thought and background. The university was a breeding ground for interdisciplinary research, and a female dean led the computer science college, which was made up of 30% women.

“I never felt like an outlier as there,” Sun noted.

A Guiding Hand

Sun even managed to find a female mentor at Northeastern, associate professor Tina Eliassi-Rad, who also focused on data science.

Considering that women make up less than 15% of computer-science faculty nationwide, according to the National Center for Education Statistics, Sun said that having a female mentor was especially helpful to successfully tackle the obstacles of working in a male-dominated field. “Having someone who is your senior and understands the challenges you face because of your gender is particularly beneficial,” she explained.

A particularly tough time for Sun in her career was when she was pregnant with her now two-year old son. “I noticed small differences in the way I was treated, and the first several months after he was born was really difficult.” she said.

The Politics of Data Science

Now at UCLA, Sun will teach data mining at a graduate level with plans to lead an introductory class on the topic in the future. She is currently searching for several graduate students to assist with her research in data mining.

One of her current projects is  to use social media data to estimate the political leanings of particular communities and then predict voting behavior. For example, Sun hopes she will be able to identify the ideology of a Twitter user relying on their online behaviors such as following and retweeting.

She believes this process can be more effective than the traditional techniques used to estimate voting patterns, such as public opinion surveys. “Analyzing user data on a large scale has the potential to change the way that polling is conducted,” she stated.

The Ingredients to Success

Regardless of how researchers apply data mining, Sun describes the three components necessary to “succeed” in the field. “You need the passion to identify and solve problems, the capability in math to model those problems, and the skills to turn the solutions to algorithms and code to harness the computational power ,” she said.

Looking towards the future, Sun, is working on a diversity committee started by the College of Engineering and led by Miryung Kim, an Associate Professor of Computer Science at UCLA.

The committee goals include working to bring distinguished CS speakers to UCLA as well as advocating for women in the department by way of an Association of Computing Machinery chapter created specifically for females.

But Yizhou notes that increasing the number of female computer scientists cannot be achieved solely through the efforts of universities. “Family support, mentoring,  role models, and passion for the science are all essential to fixing the gender gap in the field.”

In our new quarterly series, women from the UCLA community involved in technology, entrepreneurship, or STEM discuss their professional acheivements and their path to success. 

Data Science,
Data Mining,
Higher Ed