Standardized formats for classification metadata?

[Grant Robertson]

I am interested in including classification info in metadata. I am aware
of the Dublin Core and XMP. However, neither of these appear to specify
exactly how the classification data should be formatted within the
element.

I am interested in any standardized formats for expressing Dewey Decimal
System - DDS, Library of Congress Classification - LCC, Cutter Expansive
Classification, Universal Decimal Classification - UDC, Colon Notation
System - CNS, Bliss Bibliographic Classification Scheme, Chinese Library
Classification - CLC, and any other general or specific classification
systems you may know of.

Thank you.

 

[Joe Kesselman]

I am interested in any standardized formats for expressing Dewey Decimal
System - DDS, Library of Congress Classification - LCC, Cutter Expansive
Classification, Universal Decimal Classification - UDC, Colon Notation
System - CNS, Bliss Bibliographic Classification Scheme, Chinese Library
Classification - CLC, and any other general or specific classification
systems you may know of.

For clarification: Are you looking for formats for the classification
index (which is defined by the classification system, and generally is
not in XML syntax), or for associating documents with the indexes that
have been assigned to them by the appropriate controlling bodies, or...?

 

[Grant Robertson]

For clarification: Are you looking for formats for the classification
index (which is defined by the classification system, and generally is
not in XML syntax), or for associating documents with the indexes that
have been assigned to them by the appropriate controlling bodies, or...?

I believe the former. When you look at the LCC number on the binding of a
book you see something like:

Z
696
..A4
F88
2000

(This is a book about classification.)

However, when you put text in an element or attribute it is usually all
in a line. These five lines of text could potentially be written on one
line as:

Z696.A4F882000

or

Z:696:.A4:F88:2000

or

Z 696 .A4 F88 2000

or

Z/696/.A4/F88/2000

or any of a million different combinations. What I need is the definitive
format for expressing each different classification system's
classification codes in a single line of text within either an element or
an attribute. The actual meaning of the classification code will be
determined by the interpreting software. If there are any standards which
are designed for use within XML that would be preferred.

Thanks.

 

[Tuija Sonkkila]

If there are any standards which are designed for use within XML that
would be preferred.

Check this: http://www.loc.gov/standards/mods/

Best,
Tuija

 

[Grant Robertson]

Check this: http://www.loc.gov/standards/mods/

Thanks for the link. Unfortunately that standard doesn't specify how to
express the call number itself. I did find a link to the MARC
specifications at http://www.loc.gov/marc/bibliographic/ecbdclas.html.
Oddly, they don't specify either. However from the examples on both sites
it appears that they simply concatenate all the lines of the call number
together with no whitespace. It would be nice if they were specific about
this.

Well, that's one almost down and about 5 more to go.

Thanks.

 

[Joseph Kesselman]

What I need is the definitive
format for expressing each different classification system's
classification codes in a single line of text within either an element or
an attribute.

I would expect that to be defined by the system, or by its users, rather
than by XML per se. After all, the same issue exists if you want to
express it as a single line in a database or in any other programmatic
form, and if there's any consensus about the single-line form one would
hope it applies across these multiple programming environments.

Note that XML *can* store multi-line values, if you enclose them as
element content rather than attribute content -- so if the official form
is multiple lines, XML has no objection to handling it that way. (Or, if
you really insist on having it as attributes, you could spread the
lines/fields over multiple attributes.)

So I think the answer is "XML itself has no preference. Specific
XML-based languages may; check their specs."

 

[Grant Robertson]

So I think the answer is "XML itself has no preference. Specific
XML-based languages may; check their specs."

Thank you.

I know XML by its nature allows me to define any spec I want. What I am
looking for is the specs you mention in your last line: those defined by
the arbiters of the classification systems. I have been searching off and
on for three weeks now and can't find anything specific. Even the OCLC
specifications for MARC21 don't specifically say all the lines are
concatenated. It just appears that way in their examples. I guess I am
just looking for a definitive statement on one of the official web sites
of one of these classification consortiums. I haven't been able to find
anything other than vague generalizations. I was hoping someone here
would have a tip as where I could find what I needed.

Thanks for your help though.