Julia Stoyanovich has worked in data management in all different sectors of the tech world. During the first dotcom boom – and bust – she was at Juno, an internet service provider that Stoyanovich says had the distinction of being the first “bad” tech IPO of that time. “It made me stronger,” she says, laughing. Since then, she’s done research at Columbia University, IBM, the Max Planck Institute for Informatics, Yahoo!, and the University of Pennsylvania. Now, as an assistant professor of computer science at Drexel University, one of her passions is an initiative called “Data, Responsibly,” which promotes fairness, neutrality, and transparency in data analysis.
Stoyanovich’s diversity of experience makes her a perfect fit to be a member of the Program Committee for the 4th annual Data for Good Exchange (D4GX), to be held Sunday, September 24, 2017 at Bloomberg’s New York headquarters. The conference explores the use of data science for social good, and brings together data scientists from industry and academia with not-for-profits and government entities seeking to use modern machine learning and data science methods to address challenges in the public and non-profit sectors.
“It is very important to have communications with practitioners in the domain,” says Stoyanovich. “These are the people who inspire our work and with whom we want to share our results.” And, she says, there are different reasons to be excited about collaborating with each of them.
Companies and not-for-profits face significant barriers to entry to using data science accurately, responsibly, and well, Stoyanovich says. That’s one reason she looks forward to engaging with them. “One of our aims is to make it easier for smaller companies and not-for-profits to get the right resources,” she says.
Stoyanovich is also heartened to see government entities and policymakers actively encouraged to take part in the Data for Good Exchange. Thanks to a series of initiatives started under the Obama administration, Stoyanovich says policymakers have an increasing awareness that moral and ethical judgments are embedded into the data-driven machine learning tools we use. But when an algorithm yields an answer, it’s not at all clear how that answer was chosen over all others, and what hidden judgements were embedded in that choice. “Because we have no insight into how a decision is made, we can have very little impact,” she says. “We cannot have an informed discussion about how technology changes society if we don’t understand how it works.”
Then there are the people Stoyanovich refers to simply as “citizen-hackers,” who are on the ground examining open data and government data. “I want to understand how we can help them be more effective,” she says. “What kind of frameworks and tools would help democratize data science and make it even more accessible?”
“The audience at the Data for Good Exchange will be diverse. It will be filled with people interested in the impact they can have on society using data-driven technology,” says Stoyanovich. “If this event does not motivate you to work in data science, I am not sure what will.”