Subject: Metadata, Agent types and Generic LOs
S.J.E.Hobbs@open.ac.uk
Date: Fri 25 Feb 2000 - 16:13:43 MET
From: S.J.E.Hobbs@open.ac.uk Subject: Metadata, Agent types and Generic LOs Date: Fri, 25 Feb 2000 15:13:43 -0000
List address to send message to everyone: ifets-discuss@LISTSERV.READADP.COM
Details of current discussion: http://ifets.ieee.org/discussions/discuss.html
---------------------------------------------------------------------------
I know I'm a bit late for adding to this discussion but I have been fairly
caught up in summarising it! I suppose most interested bods have gone away
now to think about things but I'd like to add a couple of views.
Metadata:
Many contributors have commented on concepts relating to metadata and have
argued either for flexibility or for standardisation. I don't actually see
these as mutually exclusive. There is no reason why some metadata 'fields'
should not be of strictly limited and defined vocabulary (chosen from agreed
lists) and others be freely added. This approach has been successfully taken
in the ARIADNE project and in work I developed for a stock photogrpahy
agency. If this is the approach taken, the resulting system can be used by
computer systems/AI (using the restricted vocabulary items) and by human
agents (who can complement this information with the less rigourously
defined information)
I also strongly agree that metadata needs to be extensible. All things
change over time, not only the standard software tools but also approaches
to and definitions of subject area. If metadata is not allowed to evolve to
take this into account it will end up like a snake which can't grow because
it can't shed it's skin. The stock photography system (mentioned above) has
built in flexibility which allows new concepts to be added but only under
extremely controlled conditions and in a way that can be applied
retrospectively, consistently and for the most part automatically, to all
the objects already tagged.
Agent types
I disagree with Clark when he suggests that tagging is only for machines.
Given that I agree with several other contributors that, currently, only
human agents can add the necessary coherence to link objects, humans are
necessarily going to need to access and use tagging information.
Generic LOs
These have been mentioned in what I see as two different ways, firstly as
contentless 'tools' and secondly as those with content which is applicable
across contexts (Thompson). At the Open University in the UK we are
currently using the latter concept and are producing units or packages, such
as 'how to use the web to find information' and 'study skills' which are
potentially relevant and usable across contexts.
Thanks very much for a fascinating discussion.
Samantha
Samantha Hobbs
Lecturer (Digital Media)
Institute of Educational Technology
Open University, Milton Keynes
UK
Tel: +44 1908 654128
Fax: +44 1908 654173
---------------------------------------------------------
Forum website: http://ifets.ieee.org/
Forum's contact person: kinshuk@massey.ac.nz
Info on Join/Leave List: http://ifets.ieee.org/maillist.html
---------------------------------------------------------
This archive was generated by hypermail 2a24 : Fri 25 Feb 2000 - 20:43:54 MET