Hi Simon,

> But for my real interest, "metadata" should be fine by everyone. I'm 
> interested in what information needs to be attached to information in 
> a portfolio (in education, personal or professional development etc., 
> not finance etc.) The situation is that e-portfolio systems allow 
> other people to see selections of information about people and their 
> works, skills, activities, achievements, goals, statements, etc. etc. 
> A key function of effective e-portfolio systems is to allow people to 
> control access by other people to information about them and theirs. 
> For Data Protection, Freedom of Information, and intellectual property 
> purposes, it seems pretty important to allow the representation of who 
> has what rights over what information, or digital objects, as well as 
> who (friends, tutors, employers, government, ...) is given what 
> permission by the focal person.
>
> Now I am aware (though not very familiar) with work in the area of 
> Digital Rights, but I don't think that it would be appropriate to 
> invoke a fully detailed analytic schema for digital rights just for an 
> e-portfolio system. Nor is it appropriate to attach a LOM-like mass to 
> each little bit of information in an e-portfolio. Nor can one tackle 
> the whole thing as one piece: in the e-portfolio domain there are 
> inevitably mixtures of items some of which are naturally "owned" or 
> "stewarded" by the learner, some by educational institutions or other 
> bodies. It is conceivable (Scott Wilson's influence here again) that 
> some extension of Dublin Core might be crafted to fit the bill.

> Initially, I'm not expecting "the answer" (though try it on me!), but 
> I would like to know
> 1. is this a topic of interest to this list?
> 2. what approaches would people start by considering?

Don't use DRM for DP, IP and FOI. In fact, don't include any 
fine-grained access control descriptions within the record at all. If 
you must, then accompany the record with a separate access policy file 
(e.g. an XACML policy), but don't expect the target system to use it in 
the way you may expect, as most authorization processing is handled 
within application logic still, rather than by machine-readable policy.

Generally speaking access control today is managed within the hosting 
system, not read from within the data itself; if you are moving the 
portfolio as one lump into another organisation in its entirety (the 
"transition" model) then I think its unlikely that both parties would 
agree on security implementation anyway.

Metadata for handling DRM (DRELs) is also heavily patented and 
contested, so effectively innovation in this area is on hold until 
there are successful legal challenges to the existing patents. 
Otherwise, we have to license the method from the patent owner (who 
won't disclose the sums involved without requiring an NDA). XACML 
avoids this nightmare by completely separating the access policy from 
the data record itself.

For privacy, negotiation of P3P policies prior to uploading data would 
be most sensible.

> 3. what is there "out there" already?

DC is a better fit; LOM is Learning Object Metadata, and while the 
definition of a LO is so broad as to encompass the chair I'm sitting on 
writing this, I don't think its especially healthy to tag anything with 
LOM unless its definitely going to be used in something approaching a 
LO-like scenario. e-Portfolios are not really intended to be loaded 
into LMSs and used to teach classes with, so don't use LOM. (If you did 
use an example e-Portfolio as a teaching aid, then now it IS a LO, and 
you can use LOM!)

> What I'm sure of is that the best answer is quite different from the 
> best answer for learning objects. Thus I am clearly flagging that from 
> my point of view, there is no such thing as "educational" metadata, 
> but only suitable metadata for various aspects, tools, systems, 
> functions, or purposes (not clear which).

> A side-issue: the "metadata" element of IMS RDCEO (competencies, etc.) 
> sits uncomfortably with LOM in it. More suggestion that educational 
> metadata is not monolithic.
>
> Comments?
>
> Simon
>
> --
> Simon Grant, of North-West England http://www.simongrant.org/home.html
> Information Systems Strategist http://www.inst.co.uk/
> Please continue to use my established e-mail address
> a (just by itself) (at) simongrant.org