Rationale for the development of LEAP2A
In recent years, many ICT tools and systems have been built, often referred to as "e-portfolio" tools, they support several related areas of learning. They are widely used to support personal and professional development; sometimes as an integral part of courses; and often used for assessment. In portfolio-based assessment, learners use the tools to present evidence of learning, achievements and abilities. In addition, many e-portfolio tools allow their users to create presentations in support of what they want to do, including presentations to potential employers. Learners invest significant time and effort into collecting, selecting, reflecting on, and presenting the information included in their portfolios. It is likely that they will want to take with them, and use in future systems, the artefacts they have created, the reflections they have recorded, and the vital links across and between their e-portfolio information, which greatly add to the richness and usefulness of their information. It is therefore important that their efforts do not disappear, or become unusable, when they move to another institution or into the workforce. Specifications for interoperability and portability between different e-portfolio systems are crucial.
The LEAP2A specification has been developed to support interoperability between e-portfolio tools and similar systems and the portability of information between them. There are existing specifications which overlap to some extent, but those in common use have a different focus, with a great deal of complex detail that is superfluous for e-portfolio use. HR-XML, for example, deals with information held by employers about their employees, typically managed by Human Resources departments. Similarly SIF – recently aspiring to be the "Systems" and not just the "Schools" Interoperability Framework, deals with information held by schools about their pupils. Neither HR-XML nor SIF focus on the information managed by learners, whether created by them, or recognised by them as potentially useful for presenting to others.
The earliest widely recognised specification in this area was the IMS Learner Information Package (LIP), initially published by the IMS Global Learning Consortium in 2001. It presented a plausible approach to defining and representing an extended electronic CV. The specification was developed at a time when there were few relevant systems using this kind of information, indeed, before the term "e-portfolio" gained widespread currency. Little experience had been gained of real users using such systems, and LIP was not developed in close collaboration with those who were developing the systems. When developers tried it out, they found that the specification was complex, and difficult to understand or implement consistently. Few people actually adopted it, and there was little sign of any practical and effective passing of information between systems using IMS LIP. Perhaps because both the IMS ePortfolio specification (2005), and the proposed British Standard UK LeaP ("Learner Profile", drafted 2004), were built on top of IMS LIP, they also suffered from a lack of adoption. Despite this lack of practical use, IMS LIP was useful in staking out the ground.
In the UK, those involved with e-portfolio systems knew that they wanted a simpler, more straightforward specification, more in keeping with the nature of e-portfolio information. In particular, items of portfolio information are related to other items in many ways, and while every system may build these relationships into their own hierarchical structure, there is no one such structure that suits everyone. Thus, rather than IMS LIP's hierarchical model, what was needed was a relational model for portfolio information.
The development of LEAP2A
At the end of 2006, the portfolio interest group, facilitated by JISC's Centre for Educational Technology and Interoperability Standards (CETIS) agreed that we would work to developing a simpler specification. It was decided to call the new specification LEAP 2.0, because it took on the area addressed by UK LeaP, while being more in the spirit of "Web 2.0”. During the following 12 months, some of the principles of how the specification should look were developed, but how it would be implemented, or "bound", was not decided at that time.
The turning point came towards the end of 2007, when JISC agreed to provide a small amount of funding in comparison to the relatively large commitment of three initial partners, all of whom had extensive experience of portfolio system development, and of interoperability initiatives and trials. It was important that the funding was kept low – enough to help partners, but not enough to fund their work entirely. This ensured that only those who were self-motivated to pursue interoperability would want to participate. In 2008-2009, this was followed up with a larger group, including the initial three partners and four others, with some from the UK schools (14-19) and FE sector, and some from outside the UK.
A central principle of LEAP2A, upheld by the development partners, was that the specification should only cover what was currently available in e‑portfolio systems, and it should not speculate about what might be there in the future. In this way, the number of items in the specification was kept as low as practical, contributing to the ease of understanding and ease of implementation. At the same time, we were careful to keep the specification open to extension. In every case, all ideas were subject to consensus approval by all the project partners. One of the early points of consensus was that we would build the specification on top of the Atom Syndication Format, which is very widely used, and itself easily extensible.
What are the results so far?
At the time of writing (April 2009) a first stable version of the specification has been fully drafted and is now fixed, apart from corrections. Because of the way the project has been set up, all partners are already well into implementation of export and import processes using the specification.
The final "A" in LEAP2A refers to Atom. The Atom Syndication Format was deemed suitable as it is designed for representing blog feeds, and portfolio items are generally things written by learners, just as their blog entries are. Often, e-portfolio systems include some kind of blogging facility, and those entries can therefore immediately be represented in LEAP2A. Other portfolio items are represented in LEAP2A in a similar way to blog entries, though each item is also given a type, to distinguish between the vitally different kinds of information seen in an e-portfolio system. Apart from the default "entry" type, suitable for blog-like entries, the initial version of LEAP2A has these types:
The development of LEAP2A continues to be completely open, and anyone can see more detail through the CETIS wiki (or just search for "LEAP2A").
Relationships between items of portfolio information are represented as Atom links. Beyond the links already defined in Atom, LEAP2A specifies eight pairs of directed relationships, one of the pair being one way round (like "has evidence") and the other representing the same relationship the other way round (like "is evidence of").
Dates relevant to the items are also specified explicitly in LEAP2A, as these are vital to the tools which manage and display information, for example, about past activities. LEAP2A dates are additional to the dates of publication or update specified in Atom. LEAP2A also represents other information commonly attached to portfolio items, and amongst this, we also decided to specify an optional small set of "personal data" so that names, addresses, contact details etc. could easily be included in LEAP2A exports. Personal (and organisational) data is treated only as subsidiary to main portfolio items, as they do not by themselves belong to the core "territory" of e-portfolio information. Personal data was only specified in LEAP2A after careful evaluation of other options: none of the existing specifications (including HR-XML and SIF) covered just what our e-portfolio systems required.
Partner systems implementing LEAP2A now include:
- general e‑portfolio or similar tools used for PDP, including PebblePad and Mahara;
- specific e‑portfolio tools used in Higher Education, such as Newcastle University's ePET;
- portfolio systems used in the UK schools (age 14-19) sector, such as Passportfolio;
- one European system with a foothold into business HR systems.
What does the future hold?
The essence of success for an interoperability specification is to persuade as many systems as possible to use it. We feel we have a very solid basis for this, but naturally we hope that other systems will be motivated to join our consensus. Systems that would be very welcome, and where mutual benefit would be expected, include:
- e-portfolio systems widely used in particular disciplines, e.g. medical education;
- any systems using blogging in an educational or e‑portfolio context;
- European and North American e-portfolio systems, including the Open Source Portfolio;
- systems implementing electronic versions of the Europass mobility documents - the European Europass Diploma Supplement, Europass CV, etc.;
- systems planning to implement the UK's future Higher Education Achievement Report.
Systems in use in particular disciplines may have extensive frameworks of skill or competence, and part of the job of the system is often to allow learners to track their own learning and development against skill or competence objectives. One of the most salient items on the LEAP2A agenda is to build on the start made by LEAP2A, towards an agreed way of representing such frameworks in a way that is easily related to individual portfolios, without needing to represent whole frameworks within individual e‑portfolios.
Some of those involved in the development of the IMS ePortfolio specification have agreed that another very useful development would be to allow similar information to LEAP2A to be represented in a directly viewable document on the web. RDFa technology (see e.g. http://www.w3.org/TR/xhtml-rdfa-primer/) promises just that within XHTML. In order to facilitate this, another "binding" of the same LEAP2A concepts will need to be built. Due to the way in which LEAP2A has been constructed, this should be possible. This would effectively enable LEAP2A for the Semantic Web.
One of the most interesting, but least technological, challenges in the near future will be to create effective governance for LEAP2A. Different standards and industry bodies all have their own governance structures, but it is widely recognised that these are not appropriate for parts of a specification that are under development, or which require versions to be updated in good time. We intend to create a structure that allows those who have invested in implementing LEAP2A a clear voice and influence in steering and deciding its future. Other e‑portfolio and related systems vendors and developers are invited to join us, confident that we will continue to work towards the ultimate benefit of the users of all such systems.
Thanks to Shane Sutherland and all the other Portfolio InterOperability Project partners, and to JISC and CETIS colleagues.
Learning Technology Advisor for e‑portfolios, JISC CETIS