A bomb explodes on the East Coast, killing dozens of people. Law enforcement is quickly dispatched, but behind the scenes, the U.S. Bomb Data Center is quickly gathering data: what kind of bomb was used, who might have been involved, when the attack occurred, and so on.
Intelligent Software Solutions, a Colorado company that works for clients like the Department of Defense and the National Intelligence Community Agencies, typically builds software for bomb data centers that captures data and integrates it with information from other repositories.
But while government work has sustained the 16-year-old company, Kent Bimson, its chief scientist, is now exploring another application for ISS’s counter-terrorism software: preventing kids from dropping out of college.
In his spare time, Bimson is a research professor at the University of New Mexico, a school that is working to improve student retention rates by developing predictive metrics based on student data. About a year ago, Bimson started talking with the university about whether ISS’s government technology could help identify the students most likely to drop out. Since then, ISS has begun conversations with other universities, including Auburn University and Colorado State University, though no contracts are final yet.
With students, “there’s a lot of historical data that comes from different sources, that might be in different formats. It could be text, it could be structured data. Our technology is perfectly suited for that kind of problem,” says Binson.
He describes ISS’s potential software application for universities as a layer cake. The bottom layer is a set of services and modules that allows administrators to access legacy data (different legacy databases that weren’t designed to work together). The second layer is the glue–the layer that integrates data from all the different legacy sources. The third layer is the application that’s built on top of the newly integrated data, providing ways to display student retention data, and allowing people to query that data easily. The fourth and final layer consists of specialized analytics.
The University of New Mexico, for example, is building a connection graph of university courses that looks at how the courses are all related to one another. Another school might be interested in building analytics around the social connectedness of students and retention rates. “We’re not so interested in building a specific solution as building a framework that allows universities to build their own solutions very quickly,” says Binson.
While ISS doesn’t have a software ready for universities quite yet, Benson says that it will ultimately be cheaper than student analytics software currently in use. If it makes financial sense, Binson would also be open to working with high schools. But first, ISS needs to make sure that it has a viable market in higher education institutions.