The Open Source Portfolio 2.5 should:
- Add goal awareness to portfolio tool. It should be possible for each page within a freeform portfolio to be associated with one or more goals.
- Express goals in a semantic web format in the XHTML of each page of a portfolio. This would allow for the existing export capability to include semantic data. The format could be RDFa (a method for embedding semantic web data within XHTML that is slated to become part of the XHTML 2.0 standard in 2010) or a microformat (a de facto standard for semantic data in existing Web 2.0 software).
The first of these requirements was discussed in Ann Arbor and generate significant enthusiasm, at least within the group discussing the freeform portfolio. Taking the next step of including references to the goals in the XHTML of the portfolios the Portfolio tool produces-which is a new suggestion-is likely to require only minor alterations to the tool. An RDFa implementation would require the ability to reference each goal as a unique URL, and this might require some new development in the Goal Aware Tools (and might also have implications for the community library).
I think this proposal is worth community considerations for three reasons. First, it's a way of joining up a number of requirements and strategic priorities discussed in Ann Arbor. It connects community interests in goal awareness and freeform portfolios, two key areas of needs and energy within the community. These two tools likely to already be a focus of work in 2.5
Second, and probably most importantly, having the ability to extract data about goals appropriate for a computer audience embedded within a much more flexible visual and navigational structure of text and images designed for human consumption would enable cross-institutional assessment and research initiatives. With this capability, the participating institutions could address calls for increased accountability in a manner that doesn't impose reductive restraints on the portfolio learning experiences of students. Patterns in large number of portfolios could be quantified and analyzed, but the portfolio activities through which this data is gathered could still encourage the creativity and more associative thinking powerful for identity development. This seems both crucially important at this point in the history of higher education and eminently fundable. Seeking outside funding was another priority shared by many of the Summit participants.
Third, moving in this direction would help the Open Source Portfolio regain the leadership position on ePortfolio interoperability to which we've aspired over the years but never achieved. This approach avoids much of the complexity we encountered when trying to figure out how to implement the existing IMS ePortfolio Specification. While the results, at least in the short term, would not be compliant with this specification, they would be directly in line with the trajectory on which ongoing and planned ePortfolio standards work is headed. I am working with colleages in the UK and Europe to begin a new version of the IMS specification that will center on RDF embedded in XHTML, probably via RDFa in the long term, but possibly via microformats initially. The already commenced LEAP 2.0 project, a British standard, will follow this approach, as might some emergent work about competency frameworks through ISO and IEEE. By getting out a bit ahead of this work, we can drive the standards work to match the reality of our experiences rather than react to it. We can likely emerge as the reference implementation as we had hoped to be for IMS ePortfolio. Because this approach is much easier to implement than the current IMS specification, I think it's likely lots of other vendors with buy in, giving us a more meaningful payoff for our work more quickly.
Technically, this approach makes sense for a couple of reasons. The advantages of embedding semantic data within XHTML have already been mentioned: It doesn't impose on human-readable structure and it's relatively easy to implement. The semantic web approach add new strengths. First, it puts the focus squarely on relationships between data without requiring a central authority to define and mediate those relationships. One of the challenges we faced in trying to implement existing standards is that these seemed to require agreement about a central set of data types or a common vocabulary across institutions, and that looked nearly impossible to achieve. A semantic web approach allows each institution (or even individual) to do things their own way, but also allows for the relationships between different terms, fields, standards, etc. to be defined over time, again without the need for central coordination. Unlike when we began the IMS ePortfolio work, there are now very good tools and code libraries for working with RDF data, including for extracting RDF from XHTML.
If this direction is one to which we decide to commit, after OSP 2.5, next steps might be to:
- Import externally generated XHTML that (perhaps initially in the form of a ZIPed Portfolio) and generate goals / goals associations from it for use within Sakai goal aware tools
- Allow tagging based on IMS ePortfolio PortfolioPart types, such as product, achievement, reflection, etc.
- Express other kinds of associations this way (such as those from Matrix) in exportable form
For an introduction to RDFa, see: http://www.w3.org/TR/xhtml-rdfa-primer/
For more in microformats: http://microformats.org/about/