SyntaxHighlighter

Saturday, September 10, 2011

Feeds of Flickr Photos Depicting Pleiades Places

Some months ago, ISAW started adding Pleiades machine tags to the Ancient World Image Bank (AWIB) photos we've been uploading to Flickr. This post will explain what that means, how it might be useful to you and how you can add Pleiades machine tags to your own photos so we can find out about them.

Updated: 8:45pm EDT, 10 September 2011 (changes highlighted in orange).
Updated: 10:43am EST, 20 December 2011 (some of what's here is now superseded by recent developments; see further this new post: Pleiades, Flickr, and the Ancient World Image Bank)

Pleiades Machine Tags

Pleiades is a collaborative, open-access digital gazetteer for the ancient world. AWIB is an open-access publication that uses the Flickr photo-sharing site to publish free, reusable photos of ancient sites and artifacts. Machine tags are an extension to Flickr's basic tag-this-photo functionality that "use a special syntax to define extra information about a tag" (Aaron Straup Cope, "Ladies and Gentlemen: Machine Tags," 24 January 2007).

A Pleiades machine tag looks like this:
pleiades:place=795868
where 795868 is the stable identifier portion of a Pleiades Uniform Resource Identifier (URI).  In this example, the URI corresponding to the machine tag above is:
http://pleiades.stoa.org/places/795868
Note what's in common between the machine tag and the URI (highlighted in yellow).

What Pleiades Machine Tags Are Good For

The Flickr API makes it possible to request lists of machine-tagged photos in the RSS webfeed format. So, to get a list of all photos in Flickr that are tagged with the example machine tag above, pop this into your feed reader:
http://api.flickr.com/services/feeds/photos_public.gne?tags=pleiades%3Aplace=795868&lang=en-us&format=rss_200
The same results can be viewed in HTML in a browser by resolving the following:
http://www.flickr.com/photos/tags/pleiades:place=795868
To get a list of all photos in Flickr that are tagged with any Pleiades machine tag, try this (the API syntax supports wildcards!):
http://api.flickr.com/services/feeds/photos_public.gne?tags=pleiades%3Aplace%3D&lang=en-us&format=rss_200
The same results, viewed in HTML on the Flickr site:
http://www.flickr.com/photos/tags/pleiades:place
Feeds like these aren't just for feed readers anymore. You can add user-interface widgets to your blog or website to summarize the latest content for your readers (check out the right-hand column in this blog). You can hook up services like Networked Blogs or Twitterfeed to pass on the latest changes to your Facebook friends or Twitter followers. If you've got a web-facing numismatic database that you've already linked up with Pleiades for the mint locations, you could write custom code to pull a corresponding picture of the ancient site into your web interface (say, alongside the map you've already got).

Add Pleiades Machine Tags to Your Own Photos on Flickr

Many of you have been taking amazing photos of ancient sites and artifacts for years. Many of you have posted some of them to Flickr and shared them with great groups like Chiron, Visibile Parlare - Greek Inscriptions and Visibile Parlare - Latin Inscriptions. If you'd like these photos to appear in queries and feeds (like those described above), right alongside the photos that we're publishing via AWIB, all you have to do is add the appropriate Pleiades machine tags in Flickr. Just look up your site on Pleiades, copy the numeric ID from the URI in your browser's location bar, append it to the string "pleiades:place=" and tag your Flickr photos with it. In this way, you can help us improve findability of good photos of ancient sites and the artifacts found there for everyone on the web. Who knows ... maybe enough people will join us in this effort that we can someday get the Flickr development team to give Pleiades machine tags some extra love.

Kudos to:

8 comments:

Tom Elliott said...
This comment has been removed by the author.
Tom Elliott said...

A huge hurrah! for Dan Diffendale, who has already begun adding Pleiades machine tags to his spectacular collection of photos on Flickr.

John Muccigrosso said...

Seems like all the Open Street Map - linked images also have "Additional Info" tags that say that "X is a feature in Open Street Map". I'm assuming that's somehow automated.

Tom Elliott said...

John: the Flickr development team from time-to-time select certain machine tag name spaces on which to bestow some "extra love," i.e., they create custom behaviors in the Flickr user interface around those namespaces. Open Street Map is one of them. You can read more about it in a couple of Flickr developer blog posts: here and here.

Tom Elliott said...

With Dan Diffendale's permission, I'm posting a series of email messages here that I think are of potential interest to other readers of this post.

-------- Original Message --------
Subject: Pleiades machine tags and your photos
Date: Sat, 10 Sep 2011 18:22:41 -0400
From: Tom Elliott
To: Dan Diffendale

Dan:

I noticed that you started using Pleiades machine tags this afternoon. Thanks so much! I've long been a fan of your photography. I hope we'll eventually be able to use this mechanism to start building automatic links on the place pages in Pleiades. By adding those machine tags to your photos you're going to be giving Pleiades users a much richer experience when we roll out that functionality.

Do let me know if you have any questions or suggestions, or if you have any critiques of our online publications and resources. Also, if you run across any missing places in Pleiades, drop me a line.

Best,
Tom

Tom Elliott said...

-------- Original Message --------
Subject: Re: Pleiades machine tags and your photos
Date: Sat, 10 Sep 2011 20:14:37 -0400
From: Dan Diffendale
To: Tom Elliott

Hi Tom,

My pleasure. The linking has a lot of potential; I'm looking forward to its further development.

One question that occurs to me is what the minimum standard for artifact provenance is or ought to be; is it worth machine-tagging objects "probably" or "said to be from"...? Also, in a few cases, I've added two machine tags, one for the archaeological provenance and one for the object's place of production--any thoughts on that? I'll keep a running tally of missing places; I've noticed a few so far, mostly Bronze Age sites--not surprising, given the Barrington's focus. I
haven't seen an explicitly-stated chronological range for Pleiades; are you still working within the 1000 BCE-650 CE limits?

Best,
Dan

Tom Elliott said...

-------- Original Message --------
Subject: Re: Pleiades machine tags and your photos
Date: Sun, 11 Sep 2011 15:10:12 -0400
From: Tom Elliott
To: Dan Diffendale

FWIW, I tried to create a quick-and-dirty slideshow of everything with Pleiades tags on our main AWIB page at http://isaw.nyu.edu/awib, but it turns out to require more intervention in the page templates than I'm spun up on, so I'm going to have to put it in our web developer's queue behind some [other work].

But meantime we can look at the Pleiades interactions, which is really the core issue at the moment. Though the long-term vision is to find ways to make AWIB, or something AWIB-like, a collaborative venture in which ISAW plays the role of cheerleader and problem solver and offers participants long-term image preservation services through the medium of the NYU Faculty Digital Archive (in case Flickr self-destructs, etc).

I think the "probably" and "said to be from" options are fine. I wonder thought whether we ought to introduce a new tag or two that would let us differentiate between pictures *of* a site and pictures of objects *from* a site. Maybe something like:

pleiades:place=12345 means "picture depicts the site (or some built component thereof)" ... or maybe (in light of the following) something like pleiades:depicts=12345

pleiades:findspot=12345 means "picture depicts object found on the site"

pleiades:origin=12345 means "object originating on the site"

These terms would mirror the vocabulary we've been toying around with for a long time regarding semantic web stuff:

http://gawd.atlantides.org/terms/

What do you (and other addressees on this message) think of that?

It would be great to add the bronze age sites you're finding to Pleiades. Can I set you up with a username and password and we can then talk about the quickest/easiest way for you to make those suggestions? Pleiades is expanding beyond the boundaries of the Barrington, both in temporal and spatial terms. We've already pushed much further back for a few sites in Egypt, and forward in time for Byzantium, and there's a group of folks in California who've just got an NEH startup grant to start adding Ancient Near East places to Pleiades. [...]

Best,
Tom

Tom Elliott said...

-------- Original Message --------
Subject: Re: Pleiades machine tags and your photos
Date: Sun, 11 Sep 2011 21:08:41 -0400
From: Dan Diffendale
To: Tom Elliott

Tom (et alii),

depicts, findspot, and origin all make good sense; pleiades:attestsTo could also be used when appropriate, e.g. the Brea Foundation decree:

http://www.flickr.com/photos/dandiffendale/3035342766

or the kouros from Phigaleia:

http://www.flickr.com/photos/dandiffendale/5252231681

pleiades:observedAt could potentially be employed, though perhaps less useful; most (but not all) of the examples I can think of were observed somewhere else within the area already covered by pleiades:findspot=12345.

I'm certainly happy to get set up with an account, but I wonder about the long-term merits of adding sites piecemeal according to whether or not I (or anyone else) have a relevant photo. Coming down as it does
to a question of priorities and resources, perhaps it's the best option for now?

In terms of the (Greek) Bronze Age, Hope Simpson and Dickinson's Gazeteer of Aegean Civilisation in the Bronze Age springs to mind. The Gazetteer seems to have been used as a source by the ArchAtlas team:

http://www.archatlas.dept.shef.ac.uk/atlas/atlas_sources.php?source=gacba&detail=yes

but only the major sites (generally those already covered by the Barrington/Pleiades) are available at this point. Cracking into the Gazetteer raises the issue of significance; some entries record the findspots of sherd concentrations--and we come rapidly into survey territory: how does Pleiades define a "place"?

cheers,
Dan