Tom Elliot | Central Unit blogs

Pleiades project at UNC

Sorry for the paucity of writing recently, I’ve had some time out to get married and have a honeymoon diving in the Maldives. There’s been quite a lot going on recently and I’m still catching up slowly.

Pleiades image I’m not sure how many people are fully aware of the fantastic work that is going on at the University of North Carolina (Chapel Hill) with regards to turning the Barrington Atlas of the Classical World into a digital goldmine based upon the Plone open-source CMS.

Organized by the Ancient World Mapping Center at the University of North Carolina at Chapel Hill, U.S.A., Pleiades brings together a global community of scholars, students and enthusiasts to expand and enhance continually the information originally brought together by the Classical Atlas Project (1988-2000) to support the publication of the Barrington Atlas of the Greek and Roman World (R.J.A. Talbert, ed., Princeton, 2000).

Our name, “Pleiades” (the daughters of Atlas in Greek Mythology) reflects both this heritage and the forward-looking goal of collaborative diversification.

I had the pleasure of collaborating with Tom Elliot and others on the Digital Coins project last year, and I’m currently a technical observer of their project. They have introduced a huge array of features, some of which have inspired me to add features to a couple of my projects (Sean Gillies’ post on geoRSS for instance).

The Pleiades project has just released a large amount of their information into public circulation with permalinks for use within software that support simple geoRSS specs. They offer several of their grid squares, and the current release of all places in the Barrington Atlas under a Creative Commons Attribution-Share Alike 3.0 License. Awesome! Keep it up chaps.

Barrington Atlas in Google Earth
All places viewed via Google Earth

View Larger Map
All places viewed via Google Maps interface

Sean is really active at pushing geographic developments and loves his Python. If you want to get involved, he has a Google group on RESTful geographic technology and chucks some thought provoking ideas around at his blog (Tom has just started his own blog as well). I’m monitoring what they are up to as I’m just about to begin specifying the next stage of development for the Scheme’s database. The Scheme’s database is about to hit 200,000 records and 300,000 objects recorded, but it has inherent flaws and functional problems that were left unresolved by Oxford ArchDigital before they liquidated and were bought out by Clarinet. We’re taking the development work entirely in house now and I am proposing to create a bespoke software solution on either a RAILs or PHP codebase, with a RESTful API, with the www.findsdatabase.org.uk URL becoming obsolete (page rank and linking at the moment is not an issue!) and the new dataset residing at www.finds.org.uk/database with friendly permalinked URLs. I want to combine the Treasure Act system (currently in a closed .NET application within the British Museum), the Celtic Coin Index data, possibly Peter Guest’s Roman coin project (subject to agreement) and the Scheme’s data into one database and then have the facility for geographic extensions to the system and the ability for external developers to reuse our data and produce fantastic mashups. Several of the brilliant things that the Pleiades project offer I want to emulate with the http://www.unc.edu/awmc/pleiades/bibliography/ – Zotero extension ready. The Scheme has over 10,000 references that could be cited by academics easily.)

Hopefully over time, institutions around the world will start to catch on and see the Pleiades project and others that get on the wagon as innovators. We shall see. Read Tom’s latest post on his blog and see whether you think he’s heading the right way!

Mush! Sphere of influence using CCI data

Last week, Sean Gillies, the lead developer for the Pleiades project at University of North Caroline blogged about his experimentation with an application he is building called Mush. He’s taken 2 georss feeds and has combined them using a PULL method to determine the sphere of influence. He writes more about this at Import Cartography and specifically uses information drawn from the Celtic Coin Index to represent his example. Fantastic and just what I wanted people to start doing with the data that I’ve bent.

As Tom Elliot put it very eloquently in his email to the Pleiades project mailing list:

Sean’s demo application takes

(1) an XML webfeed (a query result, in fact) from http://www.planningalerts.com (a site that searches “as many local authority planning websites as it can find” in the UK and provides details of development applications in user-specified locales)

and (2) a separate XML webfeed from the Oxford Celtic Coin Index (via the Portable Antiquities Scheme; kudos to Dan for implementing that feed as part of the coin index’s API!)

It then parses each for georss tags (i.e., encoded spatial information), then puts 1km radius buffers around each encoded point and then looks for intersections between any of those circles. The resulting intersections are provided as another, new feed, which can also be pumped directly into an online mapping tool (in this example, Sean runs it through google maps).

Result: a visualization (or shareable information) about development applications affecting areas where Celtic coins are known to have been found.

The method, of course, is extensible to almost any combination of feeds that have geo information in them.

Imagine what could do automatically and on-the-fly with similar calculations on 1km bubbles around pleiades places and similar bubbles
around:

* epigraphic findspots GPS-gathered by the Inscriptions of Roman Cyrenaica Project

* footprint records for archaeology survey projects and datasets cataloged by CGMA/MAGIS or archived/served by ADS or OpenContext

I thought that Sean’s work needed more publication over here. Good work that man. There will be loads more to come from him without a doubt.

Database interfaces, xml etc

I’ve just been to a 2 day seminar for a Numismatic Unified Database Standard (proposed by Sebastian Heath and Andrew Meadows) for the Digital Coins Network at the British Museum, which has given me some fantastic food for thought for developing web services. Some truly enlightened thinking is going on out there at the moment with regards to sharing and aggregating data from various resources to enrich work.

Dr Tom Elliot from the Pleiades/ EpiDoc project at UNC (even though he works in coffee shops in Alabama) showed the immense power of XML upon various recording methodologies. It’s something that is going to develop over the next few months/years and something I want to incorporate into my work. For example, making use of PIR numbers for Emperors and linking place names through to the Barrington Atlas (currently in Beta version). Tom’s work, I think, will have far reaching consequences for mapping the ancient world.

More information on all these will appear on http://www.finds.org.uk/DCN as we expand the information. There were some really interesting presentations by others, with Bernhard Weisser and Johan van Heesch particularly useful for me.

There’s some fantastic web resources out there at the moment, but many of these are hampered by poor user interfaces. Our own database is one such example, and something I really want to rectify. I want something like this: http://www.powerhousemuseum.com/collection/database/ I know this is a beta version, but it has some great social application technology built in (see tags and adding your own) and also the use of AJAX to dynamically search the collection. This had:

Web 2.0 large search boxes
Searches returned into same page (no page refresh)
Use of zoomify to view high res images (it’s simple, more should use this….)
Folksonomic tagging

They also run an interesting blog – http://www.powerhousemuseum.com/dmsblog/

I found their interface incredibly easy to use, the only thing I disliked was some of the styling of the pages. Well done to them.

The other database that I’ve come across recently (and as a result of the BM seminar) is the Roman Provincial Coinage website from the Ashmolean’s Heberden coin room. This can be found at: http://rpc.ashmus.ox.ac.uk/ and it has some slick features:

Add records to a purse for later use
Constant view of search criteria
Collapse side bar for extra width
Dynamic Greek unicode keyboard
Column addition/subtraction facility from tabulated view

Once we get the PAS database sorted out as a user centred interface, I’ll be a much happier chap. These data we collect should be easily accessible and used by more; the immense wealth of information has the potential to revolutionise archaeological perceptions of England and Wales.