UC Santa Cruz plans to create a Center of Excellence in Data Science Research that will develop new foundational research in data-driven discovery and decision making and provide a forum for researchers in industry and academia to exchange ideas and develop practical solutions to data science challenges.
Lise Getoor, professor of computer science and associate dean of research at UCSC's Baskin School of Engineering, presented her vision for the center--called Data, Discovery, and Decisions (D3)--at the engineering school's Spring Data Science Afternoon on Thursday, May 21. The event brought together participants from industry and academia to discuss the latest developments in data science research.
Glassbeam, a machine data analytics company, is providing initial seed funding of $50,000 to support the creation of the center and intends to become a founding member of the D3 Industrial Advisory Board.
"The gift from Glassbeam gives us some initial resources to help get the center started," Getoor said. "We look forward to recruiting additional founding members and working with our industry partners to develop new models for collaboration between industry and academia in data science."
"We are extremely pleased to help start this new center," said Kumar Malavalli, chairman and chief strategy officer of Glassbeam. "Companies have been collecting their own data for decades but solutions like ours that provide predictive and actionable analysis are only just beginning to come to market. There is so much business value that can come out of proper analysis, that finding additional solutions can't be left to individual businesses alone. It must be the shared responsibility between industry and academia. I have been involved with UC Santa Cruz for over ten years and in 2003 funded the Kumar Malavalli Endowed Chair for Storage Systems Research at the Jack Baskin School of Engineering. I know that UC Santa Cruz has the best platform and resources to drive this industry in new and significant ways."
Neutral platform
According to Getoor, D3's mission will be to develop new computational, mathematical, and statistical research in order to support discovery and decision making from complex, heterogeneous data. The center will serve as a neutral platform for industry and academia collaboration, focusing on common pain points and developing scalable data science templates that can be used for discovery and decision making in a variety of domains.
"The focus will be on the iterative process of going from data to discovery to decisions, which produces additional data that can be fed back into the process. We plan to focus especially on structured and heterogeneous data, such as the data generated by the Internet of Things, or any setting where you want to integrate disparate data from a variety of different sources," Getoor said. "That's very different from the settings that statistics and machine learning have traditionally looked at. We need to develop new mathematical models that are able to deal with not only the scale and computational issues of big data, but also with the heterogeneity of different kinds of data from different sources."
The structure of the center will be modeled after the National Science Foundation's Industry/University Cooperative Research Centers, such as the Center for Research in Storage Systems at UC Santa Cruz. Industry partners will provide input on current industry research needs through participation in research panels, data sharing, and other types of collaboration.
"It's very timely because there is a need for more collaboration between academia and industry around data science," said Getoor, who organized a panel discussion on the topic as part of the Data Science Afternoon at the Baskin School of Engineering. The panelists include Chaitanya Baru, senior advisor for data science at the National Science Foundation; Internet leader and cybersecurity expert Rod Beckstrom; Glassbeam CEO Puneet Pandit; and representatives from Adobe, Hitachi America, and Looker.
The keynote speaker at the event was Michael Jordan, professor of computer science at UC Berkeley and a leading researcher in statistical machine learning and artificial intelligence. The event also featured a discussion of data science at Netflix with CEO Reed Hastings and Caitlin Smallwood, vice president of science and algorithms.
Getoor said she and her collaborators are looking for additional founding members to kick-start the center.
