A Methodological Moiré
I’m at the point in a couple of projects at the law school where issues are starting to coalesce and cause interesting interference patterns — a methodological moiré.
In one project, I’m worried about the information architecture for the law school’s website that I’m restructuring: how much should be (and be presented as) hierarchical and how much should I feel free to web with linking strands across hierarchies; should content be used twice or only once and referenced twice if need be, etc. — and how should all of this get translated into a decently usable interface and navigation structure?
In the other project, I’m working to acquire a complex system for the law school that will support the needs of a network of scholars, so I’m knee-deep in such terms as “digital assets management system,” “digital objects repository,” “content managment system,” “course managment system,” etc. etc.
The intersection? Well, in this second project it’s become clear to me that whatever one calls the system that stores whatever one calls the thing it stores (file, object, document, data, information, and — heaven help us — knowledge), someone is going to have to pay serious attention to metadata (or classification system, or taxonomy, or indexing language etc.). A system can’t be smart before you teach it. But what should you teach it about your situation? What is our “situation” — documents, web pages, faculty, history, and so forth — such that we might make good use of the managment apparatus? One salient feature of our “situation” is that it’s scattered, legion, and cannot adequately be known without the help of a management system. And so it goes.
Part, if not most, of my difficulty, of course, is that I’m not even half-wise about these matters. But some of it at least comes from the plethora
I suppose the relation is logical and straightforward, from one point of view at least: an object managment system needs a classification system in order to manage objects; and a classification system exists to facilitate an object management system (whether that be live people getting and shelving books or a whiz-bang collocation of chipsets parsing digital data). And, I guess, like any complex enterprise there’s an element of bootstrapping required whereby you just go ahead with your best guess and repeatedly modify after the fact until the top is spinning without too much of a wobble.
One can also say that the system pair — (legal) information technology + (legal) information science — is in a state of flux, which presents real opportunities for souls cleverer than I to clarify matters and enlarge our capabilities. At the moment I’m trying to wrap my mind around topic maps, and wonder if any of you cleverer souls out there have made use of them in the field of legal research or object management. If so, or if it’s a tree not worth barking up, let me know.
It is in fact a very interesting problem, and you’re surely not the first to give it some brain time. Topic Maps and the semantic web standards RDF, OWL are indeed good tools to let you formalise that knowledge/ontology. The main advantage I see in RDF is that it flats down all metadata to the triplet subject-predicate-object allowing you to restructure metadata as you sound fit. Restructure it or present it. I see it has a tree you can grab by any leaf and make it the root of the tree, or exploration start point. OWL lets you give a «normal» tree or the one the authors meant. I’ve read only a little about topic map but I think it’s manly the same thing based more on a book index metaphor.
Here’s a project which I found very interesting, but I think it’s dead: http://www.cs.washington.edu/research/semweb/ Maybe it could help you see clear in the forest of organizing data.
That being said, there is not much I could find on the web myself. I took a graduate class for my master in CS (still working on that) that explore document managing with the semantic web. For one of my project, I tried to develop an ontology that described the relation between the artefact of the primary sources of law. It’s all in french but my work can be found here . It’s not complete and there might a lot of errors. In a nutshell, I tried to organize those artefact (statutes, regulation, judgment text) by the processes by which they come to life. I became a dad before I could finish it and well, you, know, work family and school can take most of your time… I though that my idea though could be a good one if it could be filled with data. Users could see how law evolves and take the knowledge tree form any leaf…
Ontology developpement thought is a time-consuming and costly activity. Users are used to the tools they know. I think the semantic web is suffering from unenlightement. Meaning people don’t imagine how great it could be to explore and use data, so they don’t see the need for it. I think it’s a case of bootstraping (http://www.bootstrap.org/). The semantic web will be useful when it exist, to exist people have to invest in it, people need to see the need which will be created when it exists.
This probably won’t help you much as to wether or not you should use Topic Maps, but maybe it can give you more insight. I said “maybe” ;)
You might be interested to see Prof. Ali Shiri’s recent survey of digital libraries in Canada. One of his main findings is that in spite of some good work, indexing is not generally done very well, or if it is done, it is not made useful to the searcher.
Presenting this material yesterday, he spoke on the lack of cross-disciplinary understanding in Canada (between CS, Library studies, info arch., cog. science, human-computer interface design, etc.), but the excellent collaborative work in the area in the US, gernerally under the heading KOS (knowledge Organization Systems).
See http://www.ualberta.ca/~ashiri/KOS_Project/index.html
Guillaume, Michael, thanks for your really helpful comments. I’ve already started to follow up your suggestions. And it helps to know I’m not completely out to lunch here.
This is kind of the topic, but still germane. I received a listserv message yesterday sent by Joe Luttrell of Meyer Boswell books about upcoming lectures to given on William Blackstone by Professor Wilfred Prest, who just published an edition of Blackstone’s letters for the Selden Society. Blackstone was, of course, is arguably the most famous classifier in the history of the common law. The text of the speech can be found at the Berkely webpage at
http://www.law.berkeley.edu/institutes/csls/Prest%20ch11.pdf
There is considerable material on how to organize records in various ISO standards. The ISO tries very hard to ensure consistsent use of single-meaning terms. The downside is that a lot of their stuff is just about unreadable by the uninitiated – and it is hard to know, sometimes, if initiation is worth the cost. The recent Canadian General Standards Board Standard on Electronic Records as Documetary Evidence fought some of these battles, though at a lower level (it seemed to me) than in some ISO standards. (There were team members who were active in the ISO who tried to pull up our collective socks.)
I recognize that this is not Simon’s focus, or Elizabeth’s, but it’s a field of discourse that may be related – a second cousin? I have no particular faces on the family tree to recommend.