view printer-friendly version (opens in new window)
The Open Source Portfolio Initiative releases 1.0
Earlier this year, the university of Minnesota realised that its well regarded and mature e-portfolio system would have a much better chance of reaching its full potential by open sourcing it, and getting others involved. The first fruits of that decision are now available for demo and download. Next stop: OKI and IMS "compliance"...
The whole field of software packages that can dynamically store student's academic transcripts and support personal development is booming at the moment. Just like many VLEs in their early days, e-portfolio software packages seem to be either home-grown, commercial products that are spun off home grown systems or add-ons from existing vendors.
What hasn't happened to a great degree with VLEs, however, is a number of home growing universities banding together and pooling the development effort (pace things like the UK's Bodington Common).
OSPI (Open Source Portfolio Initiative) looks like it may do just that for e-portfolio software. Though largely based on the University of Minnesota Enterprise System's electronic portfolio software, the University of Delaware and the r-smart group have co-founded the initiative, and have attracted the Carnegie Foundation, the University of Rhode Island, Georgetown University, Indiana University, and the University of Michigan. The initiative received a Mellon planning grant as well.
Possibly the most immediately attractive aspect of the OSPI system as it stands is the fact that it is pretty mature and well proven. Though the idea of a dynamic, lifelong record of a student's achievements is still new to many, some arts related departments have had similarish systems for the web for years. Minnesota's system has been developed and extended for six years, and, in its OSPI guise, is now being deployed across a variety of institutions.
Functionally, OSPI focusses on individuals: though meant to be stored on an institution's servers, students own their own portfolio, and determine who gets to see what. The system is designed to easily create any number of different, dynamic views on the whole collection of data stored in a portfolio. Access can even be time-limited, so that a potential employer, for example, has a fairly full picture during a job application process, but not before or after.
Data can be entered either by an institution's enterprise systems, or else via a pretty easy to use web interface by the learner herself. Which is automatic, and which is self-reported is determined by an institution's policies, and IT capabilities.
Technically, the system itself is based on J2EE- a popular and widely used set of technologies in Higher Education. A number of open source as well as commercial packages can either host it or integrate with it. OSPI was initially designed to integrate with the PeopleSoft student record and management information system, but the dependencies between the two have been removed for the present release. It still means that integration with that package is likely to be easiest, though.
The really big question with any such system, however, is whether and which standards are supported. This is important for any enterprise-type system, but even more so for a new class of system, and more especially one whose full potential would put it at the heart of both the administrative and the pedagogical work of a college or university. This is amplified yet further by the fact that e-portfolio systems have such a long reach across time (it is designed for lifelong learning), and across sectors (further education, higher education, the world of work and any combination of the three).
Clearly, without a standard means of integrating with other enterprise systems, an e-portfolio system is severely crippled. Without the means to output e-portfolio data in a standard format, it's next to useless.
Judging by their "Electronic Portfolios Need Standards to Survive" article in Educause (124 Kb, locked pdf) The OSPI team are extremely well aware of the issue.
The slight pickle they find themselves in, though, is the choice of standard. With the field so new, there is no clear winner among any of the proprietary formats -if there would be, they'd stand little chance-, nor is the one existing open spec (IMS Learner Information Profile) in wide use yet. Given that OSPI is an existing system with it's own proven protocols and data formats, and given that it is now developed by a federation of universities in an open source project, the temptation to simply declare OSPI the standard must be sky-high.
Yet, as MELCOE's James Dalziel (also founder of WebMCQ Pty Ltd) pointed out in "Open standards versus open source in e-learning" (48 Kb, MS Word), an open source technology is not the same as an open standard. Even if you don't agree with Dalziel's suggestion that open standards are best supported by closed source software, it is undoubtedly true that an essentially proprietary format just doesn't have the functional or political clout of a proper by-consensus specification or standard.
Luckily, the OSPI have promised that the next release will be "OKI and IMS 'compliant'". When asked, Christopher Coppola of r-smart (a company set up to work on OSPI), says that they are looking into supporting the Digital Repository (DR), Hierarchy, Assessment, Authentication and Authorization OKI Open Service Interface Definitions (OSIDs). Trent Batson of Rhode Island university adds that, slightly further along, the OSPI might be working on a specialised portfolio OSID, in cooperation with the original OKI folks.
The situation on IMS is a little less defined. Christopher indicates that IMS LIP is being looked at, but no decision about supporting it has been taken. Trent points out that the next OSPI development cycle is already too far along to wait for the new IMS ePortfolio recommendation, but they are in active consultation with the IMS lead in that group, Darren Cambridge.
A plan for a 'plug-in' data format feature is very high on the OSPI's wishlist. This would enable (a view on) the contents of an e-portfolio to be exported in a variety of standard formats. Christopher sees that function as the main avenue for supporting the work now done in the IMS e-Portfolio group.
In all, it seems that the present standards support situation is either one of transition out of a purely proprietary phase into a standard compliant one, or else a slightly uncomfortable compromise between setting their own standards and full open standard compliance. In that sense, OKI support would certainly make a big difference in system integration, but, as the OKI people will readily acknowledge, the OKI process of setting standards is not exactly open. What happens to IMS LIP and/or e-Portfolio support, then, looks like the real proof of the pudding. It would literally make all the difference in the UK.
The software, more information and a demo are all available from the Open Source Portfolio Initiative's website.