Ontology dynamics and vocabulary best practice

[Frederic V. Hessman]

> Pointer number one: an annotated bibliography of documents and web  
> pages concerned with taxonomy and thesaurus management. This is at  
> <http://www.nglis.org.uk/tipsbib.htm> (that document is Word-only,  
> but it seems they're going to put a PDF version there soon, and  
> there's an older PDF version at <http://www.govtalk.gov.uk/ 
> documents/ Bibliography2005-05-11.pdf>). This came from a chat with  
> a (paper) archivist colleague at Glasgow.

Thanks for these tips, Norman.

I flipped through a few of the entries - ugh!  All very interesting,  
but it would be nice to have someone be able to step back, glance  
around, and give us a few last (!) tips.  For instance at the  
Introductory Tutorial on Thesaurus Construction, http:// 
instruct.uwo.ca/gplis/677/thesaur/main00.htm#contents , I found the  
following

Standardizing the Form of Words

Terms collected should already be nouns or noun phrases. Here are  
some further guidelines for the form that terms should take in your  
final thesaurus.
Guidelines
Examples
Plural for things that can be counted
"TUBES"
Singular for "mass" nouns
"WOOD"
Singular for processes, properties, and conditions
"REFRIGERATION"
"WEIGHT",
"POVERTY"
Not inverted
"RADAR ANTENNAS"
(rather than "ANTENNAS, RADAR")
Excluding prepositions
"CARBOHYDRATE METABOLISM"
(rather than "METABOLISM OF CARBOHYDRATES")
Excluding punctuation marks, diacritics, special characters, and  
abbreviations
"COOPERATIVE PROGRAMS"
(rather than "CO-OPERATIVE PROGRAMS" or "COÖPERATIVE PROGRAMS")
"MUSICAL NOTES"
(rather than "(MUSICAL) NOTES" or "MUS. NOTES"

I'm not sure what a "mass noun" is - probably those with no separate  
plural (like "deer" and "information"?), but otherwise the rules  
sounds reasonable and we've basically been following them (i.e. the  
IAU thesaurus probably followed them).  Well, do we use "pre main  
sequence stars" or "premain sequence stars" or can we simply keed  
"pre-main sequence stars" (stars are "countable" so "stars" rather  
than "star")?  Like all rules, I'm sure they are also made to be  
broken....

I'll go through our list and see what things might need to be changed  
based on these rules (assuming that we can keep a few hyphens after  
all).


> The 'Art and Architecture Thesaurus' is reportedly a much-cited  
> model of good practice. It appears that archives people do indeed  
> have drummed into them the substantial costs involved in creating  
> or maintaining thesauri.

There are orders of magnitude more arts and architecture people out  
there, so we shouldn't be too bashful about just now accepting the  
idea that a thesaurus might be a good idea ;-)

On the other hand, the Art&Arch thesaurus has lots of nuances which  
we really don't want to get into (yet):
	"="	for exact equivalence
	"+/-" for inexact equivalence ("more or less")
	"<" for partial equivalence
	"+" for single-to-multiple equivalence
	"NE" for non-equivalence

or how about this section to make life complicated:

IX. Plural and Singular Term Forms

The ISO standards recommend that each language be displayed following  
its national standard even if this results in parallel displays in  
both the singular and the plural. ISO 5964 states that in general the  
singular is preferred in French and German thesauri, while the plural  
is preferred in British English and American English thesauri.  
Singular or plural use may also depend on the type of institution  
that creates the vocabulary. The Anglo-American library community in  
many cases prefers the plural for count nouns and the singular for  
non-count nouns; the museum community often applies the singular for  
both count and non-count nouns. The International Terminology Working  
Group accepts both plural and singular use in participating  
vocabularies according to the current usage of each. Equivalents are  
acceptable between singular terms in one vocabulary and plural in the  
other[17].

and they even  have an on-line term input form: http://www.chin.gc.ca/ 
Resources/Publications/Guidelines/English/Appendices/example-mtts1.html


Rick

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.ivoa.net/pipermail/semantics/attachments/20071023/e7578026/attachment-0001.html>