Comments on: SPARQL for Flickr: Picture the Possibilities http://wingerz.com/blog/2006/12/17/sparql-for-flickr-picture-the-possibilities/ life snapshots Fri, 24 Jun 2011 17:13:04 -0700 http://wordpress.org/?v=2.8.4 hourly 1 By: ~wingerz » Text indexing and query in Boca http://wingerz.com/blog/2006/12/17/sparql-for-flickr-picture-the-possibilities/comment-page-1/#comment-4671 ~wingerz » Text indexing and query in Boca Tue, 06 Feb 2007 15:15:15 +0000 http://wingerz.com/blog/2006/12/17/sparql-for-flickr-picture-the-possibilities/#comment-4671 [...] This powerful feature allows SPARQL-aware developers to roll their own APIs. It’s easy to whip up a search across the all literals for traditional text search behavior. With a little more work, you can craft more sophisticated searches, like one for authors of a paper that mentions a specific search term in the abstract (say, “march madness”). [...] [...] This powerful feature allows SPARQL-aware developers to roll their own APIs. It’s easy to whip up a search across the all literals for traditional text search behavior. With a little more work, you can craft more sophisticated searches, like one for authors of a paper that mentions a specific search term in the abstract (say, “march madness”). [...]

]]> By: ~wingerz » Thoughts on DSLR ownership http://wingerz.com/blog/2006/12/17/sparql-for-flickr-picture-the-possibilities/comment-page-1/#comment-3965 ~wingerz » Thoughts on DSLR ownership Sat, 27 Jan 2007 06:04:51 +0000 http://wingerz.com/blog/2006/12/17/sparql-for-flickr-picture-the-possibilities/#comment-3965 [...] Post-processing takes time. I’ve been shooting in RAW and tweaking each image a bit before converting to JPEG. Also, I need to do a better job of keeping track of metadata so that later on I can export to RDF and query my images with SPARQL. [...] [...] Post-processing takes time. I’ve been shooting in RAW and tweaking each image a bit before converting to JPEG. Also, I need to do a better job of keeping track of metadata so that later on I can export to RDF and query my images with SPARQL. [...]

]]> By: inkel http://wingerz.com/blog/2006/12/17/sparql-for-flickr-picture-the-possibilities/comment-page-1/#comment-2054 inkel Tue, 19 Dec 2006 18:14:37 +0000 http://wingerz.com/blog/2006/12/17/sparql-for-flickr-picture-the-possibilities/#comment-2054 I was wondering the same long time ago, and I even started to write some scripts that transform Flickr data into RDF (mostly for FOAF purposes). A SPARQL endpoint would be wonderful, lots of friends of mine love to take and search photos on Flickr, but as you've said, it's quite difficult to find proper results. In case you're interested, the rough english translation of my rants is in http://f14web.com.ar/inkel/2005/08/25/flickr2foaf.en.html I'm looking at Dan's link now, seems very interesting. I was wondering the same long time ago, and I even started to write some scripts that transform Flickr data into RDF (mostly for FOAF purposes). A SPARQL endpoint would be wonderful, lots of friends of mine love to take and search photos on Flickr, but as you’ve said, it’s quite difficult to find proper results.

In case you’re interested, the rough english translation of my rants is in http://f14web.com.ar/inkel/2005/08/25/flickr2foaf.en.html

I’m looking at Dan’s link now, seems very interesting.

]]> By: wingerz http://wingerz.com/blog/2006/12/17/sparql-for-flickr-picture-the-possibilities/comment-page-1/#comment-2003 wingerz Sun, 17 Dec 2006 21:07:39 +0000 http://wingerz.com/blog/2006/12/17/sparql-for-flickr-picture-the-possibilities/#comment-2003 Yes, I definitely agree that it would be costly and it probably wouldn't be practical to deploy a completely open SPARQL endpoint on any site. With keys you could restrict the number of requests over a given time period and track usage patterns. I'm sure it would be valuable from a business standpoint, which could drive interest (and maybe result in more optimization research). I'm sure IBM won't complain if all of a sudden there is a huge need for server cycles - will be interesting to see whether exposing this sort of functionality is cost-effective. I guess that's part of our (the SW community's) job. Will take a look at the link; it sounds very interesting. Yes, I definitely agree that it would be costly and it probably wouldn’t be practical to deploy a completely open SPARQL endpoint on any site. With keys you could restrict the number of requests over a given time period and track usage patterns. I’m sure it would be valuable from a business standpoint, which could drive interest (and maybe result in more optimization research).

I’m sure IBM won’t complain if all of a sudden there is a huge need for server cycles – will be interesting to see whether exposing this sort of functionality is cost-effective. I guess that’s part of our (the SW community’s) job.

Will take a look at the link; it sounds very interesting.

]]> By: Dan Brickley http://wingerz.com/blog/2006/12/17/sparql-for-flickr-picture-the-possibilities/comment-page-1/#comment-2000 Dan Brickley Sun, 17 Dec 2006 20:33:39 +0000 http://wingerz.com/blog/2006/12/17/sparql-for-flickr-picture-the-possibilities/#comment-2000 Exposing an unconstrained SPARQL interface could be computationally rather costly. Especially if it got heavily used. But I think you're right in general: SPARQL over such photo data would be great. I started some experiments in this direction, but based around idea of doing it against a local cache of personal or group data, ... not against the entire mega dataset. See http://spypixel.com/2006/kml/photos/about.html This is currently built by taking results from Flickr API, and reshaping them to look like the results of SPARQL queries expressed in JSON. I want to move to using the Perl CPAN Net::Flickr::Backup libraries instead, ... but they don't yet run against a group pool. Have swapped mails with the module author. If you or anyone you know would be interested to push on this idea further, it could be a fun testbed to collaborate around... Exposing an unconstrained SPARQL interface could be computationally rather costly. Especially if it got heavily used. But I think you’re right in general: SPARQL over such photo data would be great. I started some experiments in this direction, but based around idea of doing it against a local cache of personal or group data, … not against the entire mega dataset.

See http://spypixel.com/2006/kml/photos/about.html

This is currently built by taking results from Flickr API, and reshaping them to look like the results of SPARQL queries expressed in JSON. I want to move to using the Perl CPAN Net::Flickr::Backup libraries instead, … but they don’t yet run against a group pool. Have swapped mails with the module author. If you or anyone you know would be interested to push on this idea further, it could be a fun testbed to collaborate around…

]]>