Adam, Rosemary,
Firstly, thanks for the feedback. Now for the explanations of the way the
document currently looks... The document as you have it will not include
the corrections made as a result of our meeting in Stoke (should have an
updated version with you at some point today) so that questions about the
UID etc are still key concepts in the document structure. The easiest way
to deal with this is to pick off the elements one at a time as you have
both done in your previous emails:
Ground rules: Bill Olivier suggested that in the mapping exercise, we work
as far as possible within the existing elements of the IMS-LIP
specifications, and avoid using extensions unless there is no other way of
mapping and that we work within the IMS definitions of the core data
structures (eg QCL Affiliation).
Section 1 - Multiple UIDs:
I would agree with Rosemary that the idea of a globally unique identifier
is unrealistic given our current position and that any structure we suggest
should be able to cope with multiple models of identification. However I am
less sure of Adam's suggestion of using the <affiliation><affiliationID> to
carry this information where it belongs to a single organization.
We can use the fact that demographics is an element which can be used
multiple times and use it to create multiple UIDs for a single learner as
in the example below:
<identification>...
<demographics>
<comment>Institutional reference number</comment>
<contentype> <referential>
<indexid>UID_Inst</indexid>
</referential> </contentype>
<uid>960001111</uid>
</demographics>
<demographics>
<comment>HESA reference number</comment>
<contentype> <referential>
<indexid>UID_HESA</indexid>
</referential> </contentype>
<uid>111111</uid>
</demographics>
...</identification>
This would handle a simple case of more than one UID for the same
individual, and would still allow us to use <affiliationID> where we were
dealing with some form of external registration (eg General Medical
Council, English Nursing Board). This structure would keep the UID values
within the identification section of the LIP, where most systems would
expect to find them. Additionally it is in keeping with the model proposed
by the ISR-LIP mapping project in FE and is therefore more likely to fit
with what the commercial partners like BlackBoard and Fretwell Downing are
doing.
Section 2 - Involvement of more than one institution:
I agree with Adam here that his "favorite structure" <affiliation> seems
the best place to hold this information. This would enable us to cope with
situations such as the one described by Adam with Liv Hope and Liv Uni. The
awarding institution would appear in the <qcl> structure with the
institution delivering the tuition being in the <affiliation> field and the
<classification> being "member":
<qcl>...
<title>BA(Hons) Human Geography with Physical Geography</title>
<level>2:1</level>
<organization>...
<name>Liverpool University</name>
</organization>
...</qcl>
<affiliation>
<classification>Member</classification>
<affiliationID>960000111</affiliationID>
<typename>
<tysource sourcetype="imsdefault"/>
<tyvalue>Student</tyvalue>
</typename>
<comment>Student studies at Liverpool Hope University</comment>
<organization>...
<name>Liverpool Hope University</name>
</organization>
<description>LHU provides the tuition and student support for the
degree awarded by LU
</description>
...</affiliation>
...END OF PART ONE...
Sorry folks I have to go to a meeting, but I will pick this up when I get
back...
Regards,
Paul
_______________________________________________________________
| Paul Drummond, Deputy Director The Medical School |
| Faculty of Medicine Computing Centre Framlington Place |
| Tel: 0191 2436137 Fax: 2225016 Newcastle upon Tyne |
| NE2 4HH |
|_ http://numedsun.ncl.ac.uk/ __________________________________|
|