About ePortfolio standards (2) - Reflexion

I indicated in a previous post that although a number of actors are involved in the design of open standards, the mere implementation of open standards is not enough to ensure interoperability. Two systems can be based on the same 'base specification', yet be unable to exchange information. In order to solve the issue of information portability across systems using different specifications we designed for our members the CVT, a Web-service which makes it possible to translate one CV format into another format. Of course, in the translation, part of the information might be lost or demoted, in particular if the target format is semantically poorer than the source. The ability to translate data from one format into another leaves a lot of space for innovation in future standards.

How will specification and standards evolve?

I do not know in details where IMS intends to go with its current review of IMS ePortfolio specifications, but at this stage I believe that, at least for an employability ePortfolio, HR-XML specifications seem more mature as they profit from a much larger number of real-life implementations and existing certified HR solutions. On the other hand, HR-XML standards are just one element of a global interoperability framework. And such framework should take into account other standards, beyond those used in the education and academic worlds.

With emerging specifications like OpenSocial or Atom, I believe that it should be possible to design relevant interactions across heterogeneous systems. For such a framework there is a number of specifications that should be to studied:

• Social networks and social computing — FOAF (Friend Of A Friend, XFN (the micro-format version of FOAF), OpenSocial and Liberty Alliance People Service. ePortfolios are a the result of a social construction.
  
  
• Data representation — such as social graphs, mindmapping, heuristic chart and concept mapping can be extremely useful to provide meaningful eportfolio user interfaces. We need to better exploit the potential of RDF, OWL, TopicMaps, DotML, etc. as well as meta-data ( DublinCore) data aggregation (RSS and Atom) and people representation (HR-XML, IMS and hResume).

We should also take into account the issue of trust and privacy, i.e. secured access to personal data. For example, while the IMS ePortfolio framework allows the exchange of zipped packages of personal data, mainly for backup and import of whole ePortfolios into ePortfolio Management Systems, there is no real provision for sharing dynamically ePortfolio parts nor for protecting data privacy. This should be changed if we want ePortfolio take-up.

This is precisely what we are trying to address with the CV Universel (Universal CV) where the framework to exchange of ePortfolio parts is based on a HR-XML description of personal data transported on a Liberty Alliance Layer. In a next future, should also explore the opportunity to use OpenID 2 specifications.