On Tue, Jun 8, 2010 at 4:16 PM, Lee Passey <[email protected]> wrote:
> /Is/ FOAF well-established? According to Mr. Summers, the BBC is > publishing RDF which includes some FOAF so we now have one data point, > although we don't yet know who besides the BBC itself is consuming that > data. /Is/ it extremely common? Who besides the BBC consumes it? /Do/ > most agents that are looking for biographical information know how to > parse it? Which are they? Are there agents looking for biographical > information that use other vocabularies? Which are they, and which > vocabulary do they use? FOAF is well-established. I would argue that, alongside DC, it is probably *the* most commonly used vocabulary. This, however, is a difficult claim to prove. That said, when some of the most high-profile datasets in the Linked Data cloud use it, it can at least be considered "well-established": http://dbpedia.org/resource/Bob_Dylan See also: http://wiki.foaf-project.org/w/DataSources Not to mention the fact that Google and Yahoo! both have the capacity to parse and index FOAFs. > > What are the alternatives to FOAF? How common are they? Which is the > most expressive, and the most precise? > This misses the point to a degree - an agent that knows nothing about OpenLibrary can glean *some* usable data from OL and use it. No alternatives provide the ubiquity of FOAF. > Now this is probably a very small tempest in a relatively large teapot. > The fact is that it probably doesn't matter which vocabulary is chosen, > so long as the aggregated vocabulary is complete for the OL data set, > because /every/ consumer is probably going to have to do some kind of > transformation on the data before it is usable to him/her. If you look > at the "RDF" record for an OL work, you will see that it is not RDF at > all, but simply XML using the Open Library namespace. Personally, I find > that just as good as anything, and much more likely to be complete. > Actually, an RDF agent can do almost *nothing* with this because none of the properties mean anything outside the context of the OL. There is much less opportunity for reuse, which is a two-way street - when OL data is semantically identifiable, it not only can be reused by others, but also opens the door to consume and import data from other sources. > I just get a little rankled when I see people apparently jumping from > band-wagon to band-wagon without stopping to ask where the wagon is headed. Interesting you consider RDF/FOAF a "band-wagon", considering FOAF predates JSON (outside of merely being considered "a subset of JavaScript"). It might also be considered "a ranklable offense" to disparage RDF simply because you don't understand it or it's not useful to you. Do the JSON and XML serializations require this level of justification? -Ross. _______________________________________________ Ol-tech mailing list [email protected] http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech To unsubscribe from this mailing list, send email to [email protected]
