Computing and data permeate every aspect of society, from banking to health care, from retail to communications, from agriculture and manufacturing to education and tourism. We vote, drive, visit the doctor, buy a house and make our phone calls with the assumption that somebody, somewhere is collecting information, feeding it into a database, processing and using it (we hope ethically) to make decisions. Computing algorithms power services from Netflix to Google to Facebook and are at the heart of artificial intelligence (AI). Using statistics to detect patterns in massive amounts of data, AI algorithms not only predict which song you want Alexa to play next, but also help scientists monitor climate change, disease spread, population fluctuations and other global trends.
The need for tech-savvy graduates has been skyrocketing for years. Here at UW-Madison, computer science is the most popular undergraduate major, and last year, more than 3,000 students were pursuing degrees in majors offered by computer sciences, statistics and the Information School (iSchool).
High demand, along with the tantalizing prospect of fueling research and fostering growth and innovation in the state’s technology sector, prompted Chancellor Rebecca Blank to convene a working group in 2018 to determine a bold leadership strategy to meet Wisconsin’s needs and those of students and faculty.
The result was the creation, last fall, of the new School of Computer, Data & Information Sciences (CDIS) in the College of Letters & Science—the first new school in more than two decades on the UW-Madison campus.
“The new school unites the powerhouse departments of computer sciences, statistics and the iSchool under one umbrella—the College of Letters & Science—combining existing strengths to enable innovative new courses and curricula, as well as research that integrates the full breadth of L&S and UW–Madison,” says Eric Wilcots, dean of the College of Letters & Science.
The school aims to produce more graduates who have been exposed to, and trained in, computational thinking, big data, AI and related fields. It also recognizes that all students, no matter their major, need some training in these disciplines to be able to excel in the workplace and engage with computing capabilities that pose new challenges to privacy, security and the flow of information across diverse segments of society.
“The importance of our youth having opportunities in the area of tech has never been greater,” says Tom Erickson (BS, Electrical and Computer Engineering, ’80), the founding director of CDIS. “And the state of Wisconsin is positioned for innovation and entrepreneurship, especially with the university here. There’s always this conversation: How do we create jobs and get people working? The answer is staring us in the face: We educate a wider pool of students, and we help create an ecosystem that nurtures the creation of new companies.”
Tapping Leaders to Grow Talent
Erickson was tapped by Blank to co-lead the working group and eventually was selected to lead the new school. He grew up in Mondovi, Wis. (a third-generation Badger), and has been part of at least ten software companies around the world in the last four decades. He helped found Acquia (which commercialized Drupal as a platform) in 2008 and retired from the international firm in 2017 as its CEO.
“I have always been interested in how we take talent and grow it,” he says. “I want to encourage programs that allow people from any kind of competency and interest level to graduate from this university with enough tech literacy that they can find meaningful employment and leverage that knowledge for success.”
Associate Director of CDIS Kristin Eschenfelder agrees. The professor of information sciences and former director of UW-Madison’s iSchool was recruited to handle all aspects of academic programming, as well as faculty hiring and recruiting for the new school. An explicit goal, she says, is to create programs and opportunities that attract a more diverse range of students to the computing/data/information field.
“There’s a huge pool of very, very smart students who may not feel comfortable diving into computer science as a major, but are very interested in technology,” she says.
“As a society, we need all the brains we can get in this area. We have to attract new people from new audiences. And once we do that, we have to keep them.”
There is a great “value proposition,” as Erickson describes it, in the decision to locate the new school within the College of Letters & Science, home to the humanities, the social sciences, and the natural and physical sciences.
“We’ve spent the last 50 years of the Information Age making faster computers and more storage, creating environments for machine learning and harnessing big data,” he says. “The challenge we face in the next 50 years is, are we doing all of this right? What does the Information Age mean for society? That’s not just a discussion between statisticians and computer scientists. Sociologists, psychologists, philosophers, and historians should be part of it, too.”
Placing Humans at the Center
Everyone has experience with frustrating information systems that should be easy and turn out to be daunting. When technology works well, it’s freeing and empowering. When it doesn’t, it can be annoying. Even a well-designed system can be ethically questionable and even potentially harmful.
“We need people with the skills and knowledge to design systems that work better for people, that help them accomplish their tasks in the world,” says Eschenfelder. “But we also need to ask: What are the negative sides to all this technology? If students aren’t sensitized to these things, they come out the door so solutions-oriented that they just build the system without thinking through the negative repercussions of the designs they make.”
Corey Jackson, a current postdoctoral scholar who will join the iSchool faculty as an assistant professor in Fall 2021, specializes in the area of human-centered computing. He came to UW-Madison from the University of California-Berkeley, where he taught data science. Jackson says that an idea has floated around the data and computing fields for a long time: If you build it, they will come.
But that’s not always true, he says. “I’m collaborating with colleagues from IDEO [a global design and innovation company], Microsoft and Duke Health on COVID-19 research, where we’re trying to understand the challenges to adoption and use of contact tracing technologies,” says Jackson. “We’re finding that the populations most impacted by COVID-19 have the lowest adoption of those technologies.”
The resistance comes from a deeper suspicion of how their data will be used, he says. Historical cases of racism and unethical medical practices (such as the infamous Tuskegee study of untreated syphilis in African American men, which ran from 1932 to 1972) have created skepticism and mistrust within communities of color.
“The Tuskegee Study, a major violation of morals and ethical standards, was just one instance of how vulnerable communities are taken advantage of, and so it makes sense that communities of color may not be as eager to adopt new technologies,” Jackson says. “If we want to achieve higher adoption of contact tracing technologies, we need to ensure that our approach to the design of technology is intersectional, encompassing the voices and needs of people from all races, socioeconomic classes, and gender identities. This approach should permeate all facets of our thinking about designing future computing systems.”
Creating Flexible Pathways for Students
Eschenfelder intends to expand the percentage of UW-Madison undergraduates who think hard and critically about the human aspects of computing and data sciences, building “flexible curricular pathways” for students who want basic literacy.
“Right now, we have excellent options for students who want a full computer science or data science experience, but we need to increase the range of options for those who want a more hybrid experience,” Eschenfelder says. “Tech is everywhere in society. Those with a base level of literacy can be effective citizens, informed consumers, better advocates for their own health, privacy and safety.”
Remzi Arpaci-Dusseau, chair of the computer sciences department, agrees. “Many professions will require computational and data literacy, much like many professions have required a certain level of mathematical proficiency,” he says. “Our goal is to help students learn these core fundamentals, so they have the necessary foundations in place to succeed.”
For some students, he says, it will mean going deep into the material, perhaps adding computer science or data science as a second major. For others, it may be a certificate or smaller subset of courses that give grounding in core concepts. For yet others, it may be just a class or two, so as to have a slightly deeper appreciation of technologies that are shaping our world.
“We believe this approach is essential as students move into our highly digital and data-driven world,” says Arpaci-Dusseau.
Adding an Interdisciplinary Major
The data science major is an example of extending literacy across campus, while also making a data-oriented major more inclusive and broad-based. Students in this new interdisciplinary major (administered by the Department of Statistics within CDIS) will be able to apply computational, mathematical and statistical thinking to data-rich problems in a wide variety of fields. In addition to the L&S “breadth and depth” requirements in languages, literature, social sciences and natural sciences, the data science major will have electives in geography, economics, sociology, Biocore labs and other courses not typically required for a quantitative major.
“The Data Science Program Committee, led by Bret Larget [a professor of botany and statistics], did a lot of reaching out and creating partnerships across campus,” says Eschenfelder.
Eventually, Eschenfelder envisions adding requirements to the CDIS majors and certificates that pull from departments like psychology and sociology as well as departments in the humanities where data mining of texts has yielded fascinating insights for more than a decade.
“We are putting the building blocks in place for leadership in human-centered high tech,” Erickson says.
Building a Strong Faculty
Top-tier faculty are attracted to collaborative environments. The School of Computer, Data & Information Sciences hired a total of 15 new faculty this past spring, in the areas of human/computer interaction (UX), machine learning and artificial intelligence. All are research faculty from top-ranked institutions. All will help meet the demand for more courses in a broad range of areas that will attract a wider pool of students.
These faculty will be interacting with others from across campus regularly. “All three of these CDIS departments [computer sciences, statistics and iSchool] are ranked in the top 14 in the country,” says Erickson. “We’ll be leading on the research side. And those same people doing the research? They’ll be doing cutting-edge teaching, too.”
It follows, says Erickson, that students coming out of CDIS programs will be on the cutting edge of their disciplines.
Diversifying at Every Level
Accelerating recruiting and outreach strategies for underrepresented groups is a priority for CDIS, Erickson says.
“Historically, 35% of computer sciences graduates were women. Five years ago, that number went down to 18%,” he says. “Something happened, and we need to solve this both for the benefit of society and for the benefit of the industry.”
Arpaci-Dusseau acknowledges that computer sciences has work to do. Computing affects everyone; thus, it should include everyone, he says. “To make progress on these issues, we believe doubling down on recruiting is a first step,” he says. “But beyond recruiting, we want to build and maintain a strong, positive, inclusive culture. We cannot rest until all people have access to the education and research opportunities available within our department.”
Meeting goals of inclusivity in the student body can’t happen without faculty to lead the way. When female students see a woman teaching a 500-level computing class, they believe they can excel in a field inexplicably dominated, lately, by men. The same is true for students of color, whose experience with data, information and computing may not align with that of their white peers. Each of the departments in CDIS has a diversity and inclusion committee, and all of them meet regularly to improve best practices and set ambitious goals.
“We are committed to achieving better outcomes,” says Eschenfelder.
Collaborating on Research
Many CDIS faculty are working on teams connected with UW–Madison’s American Family Insurance Data Science Institute (DSI), which was created in 2019 as a campus-level center to encourage cutting-edge research and advance scientific discovery in collaboration with researchers across campus and beyond.
For example, there is a DSI COVID-19 Research Group that includes more than 100 people, including many CDIS faculty and staff, working together to interpret data and create models that help inform public health responses to the pandemic. Brian Yandell, professor of statistics, is the David R. Anderson Interim Director of the Data Science Institute.
“Data science research is all about collaboration,” says Yandell. “CDIS and DSI together increase campus-wide visibility and focus on computing and data science, drawing on the diverse strengths of UW-Madison to attract industry and foundation partnerships to address wicked problems such as transforming human society in the face of pandemic threats, depleted environments and social inequities.”
Partnering with Industry
At the heart of Erickson’s vision for the School of Computer, Data & Information Sciences is the goal of positioning the new school as a critical resource for sparking entrepreneurship and innovation in Wisconsin.
“One of the key factors that has benefited the coasts has been the creation of new companies,” says Erickson. “Ironically, Wisconsin is one of the great centers for scientific innovation. But we need to do a better job reaping the economic benefits of that.”
He cites a December 2019 report from the Brookings Institution, “The Case for Growth Centers: How to Spread Tech Innovation Across America,” that places Madison at the top of a list of 35 “potentially transformative growth centers” with promising innovation capacity, led largely by the research powerhouse that is UW-Madison.
The computer sciences department has always had strong ties to industry. Chair Arpaci-Dusseau says he and colleagues are excited to do even more to spark new companies and ventures.
“Where will new digital innovations be needed?” he asks. “How can we connect students to problems in agriculture, insurance, and other parts of our state economy, so that their innovations can help transform those fields? What elements need to be put in place so that innovators can be connected to relevant funding opportunities and professional networks, which are often critical to success?”
Erickson acknowledges that a strong foundation has been laid, but stresses that he’ll be encouraging conversations with tech entrepreneurs, venture capitalists and industry leaders around the state.
“I feel that the state of Wisconsin can be a leader in value creation,” he says. “And I believe the university has a responsibility for making that a reality.”