Well, this has taken a lot longer than it should but it’s probably time to pull my finger out and write up some thoughts on CAA and CA as promised. Fortunately Kayt has done much of the hard work of describing the sophisticated cut and thrust of post-prandial Computer Applications and Quantitative Methods in Archaeology conversation at exclusive eateries such as the Tadjiki Tearoom, 12 Apostel, and a variety of sausage stands. ArchCamp Drei has also been reported in some detail, so that just leaves me to cover the light-hearted frippery of the conference proper 🙂 Following that, a small update on the Friday of the Classical Association’s annual bash at Birmingham University. The fact I’m trying to fit this in with a half-a-dozen other things before my jaunt to Colombia means I’m going to have to write it over two posts. Don’t you just love a cliff-hanger?
As with most conferences of CAA’s size, the hardest part of being a participant is choosing which sessions to attend. This year there were 8 parallel tracks (occasionally on similar subjects, which cost me half my audience whilst presenting Crossbones :-/). Although many of the papers were interesting there wasn’t a great deal that felt groundbreaking. Nonetheless, two sessions in particular left me with lots to reflect on, viz. ‘Intelligent Knowledge Retrieval’ (chaired by Elisabeth Jerem), and ‘Asking Questions – Setting Standards’ (chaired by David Wheatley). I’ll cover each session in its own section in Part I, followed by some of the choicer tidbits from other ones and CA in Part II.
Intelligent Knowledge Retrieval
Having got the feeling feeling that Archaeology is still somewhat behind the curve when it come to data storage, with Object databases (or should that be Associative DBs?) implemented relationally, and RDF still an esoteric whisper, I wasn’t sure what to expect from this session. It turned out to be a terrific mix of projects which seemed both to break new ground whilst being disarmingly honest about their limitations.
First out the gate was HyperRecord, a REST-based data network put together by the Capitoline Museums, Bibliotheca Hertziana and a variety of other partners. The premise was relatively simple. Each data provider exposes all their records, be they text records, images, multimedia or whatever, as XML using Universal Resource Names (URNs). The record can also refer to other URNs, and so repositories can then send requests to other datastores for any other information they may hold with relevance to that URN. Because URNs don’t just refer to things, but concepts as well (such as ‘Minoan’, or ‘porcelain’), this can be a very powerful tool indeed. In fact the concern is that it could be a little too powerful. There could be some serious scalability issues if it ever becomes popular, but that wasn’t really what grabbed my attention. The great news is that the partners were putting all their stuff online in XML for free.
No, you didn’t misread that.
When we asked how they would get other organisations to do the same, Klaus Werner, presenting, was honest enough to give the expected answer – “we believe this is the way forward and we hope that others will see from it that openness is an advantage”. Let’s hope they do.
Second was an interesting content repository, TARCHNA, that combined CIDOC CRMed data with expert-written content, so that you could easily swing from factoid to overview and back again. This one specifically focussed on those wiley Etruscans but the model could be used with any reasonably focussed CH dataset. And the use of CRM suggests it might not be too hard to merge different projects to a certain degree. The biggest downside is probably the most over-laboured acronym I’ve heard since forever (Towards ARChaeological Heritage New Accessability? Give me a break…). It’s also designed to be standalone, so integration possibilities may not have been entirely thought through, but I wouldn’t want to judge that from here. Currently it looks very nice indeed.
Third, was Hans ‘Paai’ Paijman‘s really interesting work using Vector Space Modelling and Supervised Machine Learning to automatically markup semantically important words and phrases, like dates, out of Archaeological grey literature. And because it uses CRM categories, you can then search over approximately synonymous stuff (e.g. ‘Augustan’/’1 A.D’.) His stuff is specifically designed to work with documents in Dutch but again, the principle is applicable across the board. Using it to start exploring the vast back catalogue of data available in the UK could throw up some really fascinating results, including where we excavate and what time periods archaeologists tend to focus on.
Asking Questions – Setting Standards
Unfortunately I missed the first part of this session which was held on Thursday afternoon and only managed parts 2 & 3 on Friday Morning (by which time my constitution finally seemed to be adapting once again to hearty German fayre). All the papers presented were very interesting, but the following two seem particularly worthy of comment.
First of all, one of Sorin Hermon’s projects from the PIN stable got me all excited for the second time that week (the other project to be described later). The Archive Mapper for Archaeology (AMA) is a mapping tool that converts relational databases into CIDOC-CRM encoding on the fly. Sure, you’ve got to decide on the mapping yourself, but once it’s done you can handle the database in either way. This is really important because one of the biggest problems in CRM uptake is that it’s notoriously difficult for non-techies to get their head around. If you can have your cake and eat it, then things might really take off. There wasn’t a lot of fanfar around this, and there might be all sorts of problems around performance, etc. but, if it really does what it says it does, this would get my vote for the Paper-most-likely-be-a-success-and-retire-to-a-beach-house-in-Malibu.
Next, Karina Rodriguez-Echevarria rolled out another potential barnstormer. OK, so there was enough business jargon to kill even the most jaded of archaeological practitioners – but the message was in complete harmony with the Antiquist philosophy. Rather than working in isolation in the bizarre hope that our own research project/business model/heritage policy will bring us and our departments untold success (whilst making the same dumb mistakes and paying the same high price as everyone else), let’s sit down and chat once in a while. Y’know, kick around a few ideas maybe – even help each other (OK, so the last one’s crazy talk). The so-called Network of Excellence Centres (I did warn you about the jargon) are intended to do just that – bringing together CH practitioners within a regional context as well as to other local networks. Antiquist is currently in discussion to see how we we can mutually support our respective communities.
Whilst I was trawling through the accumulated flotsam and jetsam that had drifted into my blog-aggregator, a couple of pearls bobbed to the surface. You know, the kind of things that put an irrepressible smile on your face even after a 24-hour stint without your nicorette tabs. First was the following YouTubeVideo which Mia Ridge reported on. It’s some 3D magic done by one of the Max-Planck-Institutes (of which I’ve been a fan since I did some work with them on the Virtual Lightbox for Museums and Archives). Now if only I could find just a single decent photo of myself 😦