[rdfweb-dev] isbn - rdf:about or IFP

[Dirk-Willem van Gulik]

On Mar 3, 2004, at 4:08 AM, Masahide Kanzaki wrote:
> I'm considering to extract some book related information from XHTML, 
> and
> interested in how to identify a book (and thought using <Book
> rdf:about="urn:isbn:...">).
>
You propably want to leave room for other identifiers; or really fall 
back to 'any URI' is fine. ISBN in particular is a cumbersome choise 
for some fields of sicence; as ISBN numbers are a) recycled* and b) 
there is no-one keeping track of them in a consistent manner (though 
there are not-for-profit and competing commercial entities which 
compile lists  of sorts with various level of quality and consistency 
which cover a fair chunk).  At a minumun also leaving at least some 
room for ISSN numbers is good.

Dw

*: though increasingly this practice frowned upon (or outright 
banned/policed by
     the country level registrars) - so it is generally only a problem 
for older publications
     of very small publishers.