[rdfweb-dev] FOAF Query Service
Libby.Miller at bristol.ac.uk
Fri Aug 8 23:29:37 UTC 2003
This is very neat Sean!
I think when you're trying to create or check data, this is a really
handy thing to have, especially with its human-readable interface.
It would be excellent if other crawlers had getable query interfaces,
maybe using the Joseki framework perhaps?
On Fri, 8 Aug 2003, Sean B. Palmer wrote:
> Hi everyone,
> Using Libby Miller's SquishQL interface to her FOAF database , I've
> come up with a very friendly human interface that lets you do simple
> queries such as "homepage and mbox of danbri" via either a web form:-
> - FOAFQ: FOAF Query Service
> or on the command line using Python:-
> - FOAFQ Source
> The code is released under GPL 2. The syntax is pretty simple:
> something along the lines of "<foafprop> (and <foafprop>) of <id>"
> where foafprop can be any FOAF property "label" (i.e. after the
> http://xmlns.com/foaf/0.1/ bit), and id can be one of the following:
> an email address, defined as something with a "@" in it; a name,
> defined as something with a space " " in it; or a nickname.
> FOAF interface activies have recently been focussed on providing a
> brosable space for people to look through, find new relationships that
> they didn't know existed, and basically explore the social network
> that all of its participants have been working so hard to export the
> semantics of in RDF. But the other obvious use for FOAF that hasn't,
> to my satisfaction, been exploited enough is that of the simple query:
> it's commonplace to come across a new person and wonder what their
> homepage is, or weblog, or PGP key, or whatever.
> Even amongst people who've known each other for quite some time, there
> can be surprises. For example, danbri asked me today if I have a
> weblog. Now he can just go to:-
> In fact, I didn't know that danbri had a weblog (well, actually, I
> forgot); but in my random perusals of results whilst testing this
> script, I found that he does indeed have one:-
> I'd like to thank Libby Miller very much for letting me use her
> SquishQL service, and note to anyone using this interface that the
> service could fold at any time due to her box being a bit creaky. Hey,
> there are 200000 triples in there (though I've found the CGI to be
> suitably fast for this purpose). The script obeys robots.txt, so the
> plug can be pulled on it, the poor thing.
> The script does have some limitations. Firstly, it's limited to only
> those properties that are in the FOAF namespace. In fact, Libby's
> service poses no such limitations, but I've been too lazy to add the
> functionality since I don't think that anyone will really require
> it--if someone really wants this, they should try to cajole me into
> adding it. Secondly, there are various little oddities from the FOAF
> world, e.g. if you query for Aaron Swartz's homepage:-
> you'll find that there are results with a trailing slash and without.
> Duplicate names are another annoyance, too. Some people have been
> deliberately using a non-canonical version of my name just to irk me:-
> please don't forget the middle initial! :-)
> I hope that this script inspires people to do a better job than I've
> done. I'd like to see a ranking of the largest FOAF triple stores on
> the Web at the moment. I'd like each of the largest ones to have
> decent interfaces to them.
> Anyway, enjoy.
> Sean B. Palmer, <http://purl.org/net/sbp/>
> "phenomicity by the bucketful" - http://miscoranda.com/
> rdfweb-dev mailing list
> rdfweb-dev at vapours.rdfweb.org
> wiki: http://rdfweb.org/topic/FoafProject
More information about the foaf-dev