Re: Status of ANZLIC Metadata Guidelines

[qlis <qlis@citec.qld.gov.au>]

At 14:47 22/11/96 +1100, you wrote:
>Mark Wallis writes:
>
> THE METADATA PAGES CONCEPT
>> Many organisations are now realising the need to correctly document
>> their data holdings, including datasets they have acquired from other
>> organisations. Also some organisations often need to document
>> additional metadata elements other than those detailed under the
>> ANZLIC guidelines. These additional metadata elements could include
>> details relating to map projection, attributes details and storage
>> location. To accommodate this, ANZLIC has developed a `Pages Concept'
>> where ANZLIC core elements assigned `Page 0' status,  with any
>> additional metadata recorded at lower levels on `Page 1' (additional
>> jurisdictional metadata), `Page 2' (additional organisation metadata)
>> and so on. As `Page 1' elements gain acceptance throughout the
>> industry, they may be moved to `Page 0' status.
> allow organisations to combine additional Page 1 or 2 elements. This
>> will allow jurisdictions or organisations to enter any additionally
>> required metadata elements.
>
>I believe that this "pages" method is NOT the right structure for users
>to add extra fields to the metadata.
>
>I believe that the extra data would probably be related to the current
>ANZLIC categories (eg dataset, description, data currency etc) and
>therefore would be more useful if they were included in these areas.
>If the extra metadata did not fall into one of these categories then
>an extra category could be added at the end of the metadata report. 
>
>It would become too confusing to have to go down through many pages to
>find an extra field related to the same category just because it was
>not an ANZLIC field.  If people did not know what the ANZLIC fields
>were then the structure of the metadata directory would look
>disjointed.
>

I get the feeling when John says that the pages method "is NOT the right
structure" he is refering to it in a physical sense rather than the logical
sense that it was originally intended.

The intention of the pages concept was to outline a LOGICAL
model/structure/diagram (call it what you like) as to how the metadata used
to describe datasets can be linked at a national, jurisdictional (eg.
commonwealth, state, territory), organisational level.  

The pages concept DOES NOT stipulate how a jurisdiction, organisation should
PHYSICALLY go about implementing a tool (be it software based - eg. web
browser, standalone application, etc. or hardcopy based - eg. atlas,
catelogue, etc.) to enter or enquire on metadata. For example, if an
organisation wants to build a browser to enable entry and enquiry of
metadata and they want to collect additional Page 2 information that maybe
obviously related to a Page 0 element, then the tool should be built so that
the Page 2 details should be captured/displayed directly with the Page 0
details.  There is no need for the user to be displayed a warning that "You
are now entering Page 0, Page 1 or Page 2" details.  Similarly when
displaying reports, just because details are Page 2 does not mean that they
must be displayed at the end of Page 0 and 1 details.  It should be
irrelevant to the user to know whether the details they are reading are Page
0, 1, 2. 


>Many people print the reports that they find on the World Wide Web and
>the "Pages" method would mean that they have to print many pages just
>to get all the details of the report.  A full metadata report should
>not be very long.  After all it is metadata not data.  A list of
>contents could be used if the supplier felt that it was necessary.  A
>list of contents would demonstrate the illogical layout of the pages
>method.

Once again this is an implementation issue.

>If ANZLIC want people to use their metadata fields then they should
>consider how people will use the system, what they would use it for and
>the importance of the metadata that people want to provide.  The
>"pages" method would indicate that ANZLIC considers the ANZLIC fields
>to be the most important fields to all users where in reality they are
>not.  Especially in themetic directories.
>
>It IS important that the ANZLIC fields be incorporated in data
>directories to provide conformity but to consider that the ANZLIC
>fields are the most important, and therefore should be at the very
>front of each metadata set, is arrogant and is not considering the
>potential users of the system.  Thus they may find that people will not
>adopt this system if it dicates the structure rather than provide a
>standard.

I think it is a bit off the mark to state that ANZLIC fields (Page 0) "is
arrogant and not considering the potential users of the system".  The ANZLIC
Metadata Guidelines that describe Page 0 level metadata fields were produced
after 2 rounds of industry wide consultation in the form of discussion
papers.  Much feedback was obtained from this consultation process and was
included into the final document. Page 0 fields are no more important then
Page 1 or 2 fields. What IS important is that the land information community
starts/continues to collect metadata as one community and not dozens of
separate communities as in the past.


Regards.....Lindsay Hegarty

Department of Natural Resources
Land Information Management Subprogram
Locked Bag 40
Coorparoo DC
Queensland 4151
Australia

Phone: 61 7 38963517
Fax:   61 7 38963433
Email: qlis@citec.qld.gov.au