Fifty words or so is not a lot of space to pitch an idea, but the apparent simplicity of this format really does concentrate the mind. My entry to the TSO OpenUp competition was simply this:
“I would like to be able to programmatically access data used to power the UCAS course search. As universities open up more data relating to courses (from library usage data, to open educational resources tagged by course code), access to UCAS data would make it easier to create aggregation and distinctiveness services that span multiple institutions offering similar courses.”
But what does that actually mean?
UCAS (Universities and Colleges Admissions Service) is the organisation set up by the UK’s universities to administer applications for first-time undergraduate degrees. Each year, universities submit details of the courses they are offering, along with search keywords, to UCAS. In turn, UCAS maintains a ‘universal catalogue’ of offerings, as well as the formal application route, on behalf of the UK’s Higher Education sector. The only way of accessing this complete set of data is by searching on the UCAS website.
By opening up the UCAS course catalogue, we can provide a skeleton for a whole raft of independently published services that can be built on top of that data. By making a complete course catalogue (along with course codes) available, service providers can build services relating to course offerings that span the UK Higher Education sector, safe in the knowledge that they are not creating bias through omission. This is one problem that occurs if we are trying to crowd source the data, we may get complete coverage of courses offered by one university, but sparse coverage of those offered by another. If anything, this may lead to more efficient applications to university through UCAS because of more informed choices being made by potential students.
The first challenge we face is getting hold of the data. The simplest way is for UCAS to make it available. A slightly less convenient route is to request the data from each university, but this runs the risk of getting the data back in a variety of possibly inconsistent formats. Another approach might be to pick up the data from each university in a standardised way, an approach that JISC have been exploring through various projects relating to the open publication of marketing information for courses using the XCRI Course Advertising Profile (XCRI-CAP) data format.
Once the data skeleton is in place, more data can be aggregated around it. There is a wealth of statistics published around higher education each year: from completion rates and qualification rates, to the national student satisfaction survey; from skills maps, to teaching quality scores and employment rates after leaving university. There is also data relating more generally to the universities themselves: student/staffing ratios, research quality, even the state of repair of the physical buildings. At a more detailed level are descriptions of modules that can be taken as part of a particular degree course, along with syllabi, example course resources, and even related library loans data.
With the enriched data available, third parties will then be able to build services on top of that data. Improved course discovery and comparison services offering a far richer or personalised experience than the one provided by UCAS spring immediately to mind. TSO may choose to develop one or more services on top of this data, or publish the data as a service for third party developers. Or both.
Either way, the first step is this: to open up the data.
Tony Hirst is a lecturer in the Department of Communication and Systems at The Open University. A copy of the pitch he made to the OpenUp competition panel is available on his blog: TSO OpenUp Competition – Opening Up UCAS Data