In message <[email protected]>, Karen Coyle <[email protected]> writes > >I do not know if they've been keeping FAST up to date -- OCLC research >projects tend to fall into black holes. If you are interested in using >it you should contact someone on the project.
Have done so (though mail to the named project leader bounced - never a good sign!). We'll see what they have to say. >What this means is that if you take the 393,+ terms from LCSH, they >will match SOME subject headings in bibliographic records completely, >and will provide a left-anchored match for other headings, and will >NOT match some headings where it was valid to insert an intervening >subheading based on the rules. (The actual number of individual >subject headings in library records is in the many millions, probably >at least 100 million). > >Aren't you glad you asked?! Actually, I am. And thank you for that concise and helpful summary of 30+ years of cataloguing practice! From a Linked Data perspective, my interest is in what value the URLs minted for LCSH might have, and I have a better sense of that value now. In general, the challenge for LD is that nearly all of the data "out there" which we might want to link is expressed as strings, not as URLs. And strings are "false friends", because their meaning is clear to each human reader, but the meanings which are taken will vary. So it's good to find _any_ area of human activity, like LCSH, where you can be reasonably confident that all instances of the same string have the same intended meaning. It sounds as though a web service which took instances of LCSH headings and - where there is an exact or partial (left-anchored) match - returned the corresponding URL would allow a proportion of LCSH headings for existing catalogue entries to be expressed as Linked Data. However, that proportion might be quite low. And, as you yourself have blogged [1], the real semantic value is locked into the components of a heading. So the next step on this journey requires the authorities behind LCSH to be released as Linked Data as well. Then we will be in a position to assess how feasible it is to allocate URLs, and hence shared understanding of meaning, to the components of LCSH headings such as "Civil War, 1861-1865". Richard [1] http://kcoyle.blogspot.com/2009/05/lcsh-as-linked-data-beyond-dash-dash.h tml -- Richard Light _______________________________________________ Ol-discuss mailing list [email protected] http://mail.archive.org/cgi-bin/mailman/listinfo/ol-discuss To unsubscribe from this mailing list, send email to [email protected]
