[Sigia-l] Controlled vocabularies for facet work

[Travis Wilson]

At Fri Jul 12 2002 03:16:24 EDT, Maryanne Rosenberg wrote:
> > Maybe the new and exciting thing here is that faceted
> > classification is being exposed to the user via
> > navigation design.
>
>Good point. There certainly is an opportunity for interface design
>innovation in exploiting metadata to build more powerful navigation
>(rather than search-based or single-hierarchy) interfaces.

This is what I've been working towards for a while -- to present the power 
to the users without scaring them off. I think the first step of that 
opportunity is to give them something familiar, e.g. a menu that looks like 
a single hierarchy but which spans many facets. Once the user is used to 
that, don't you think it'd be easier to explicitly introduce facets? My 
work now is creating the data model and the software to let that happen, so 
that others may design the interfaces. See http://facetmap.com for a 
concept demo that explains it better.

I've always believed that the true power of FC is that, contrary to 
intuition, items don't have to "live" anywhere on a site. It solves the 
"buddha nature" problem that Karl covered a few months ago 
<http://www.info-arch.org/lists/sigia-l/0202/0049.html>. Most of us are 
aware by now that a page or resource can be in many containers at the same 
time, and that good search interface can exploit this complexity, but a 
user may not be. The idea of my Yahoo-style FacetMap interface is that the 
user is still allowed to think that an object has a buddha nature, even 
though it doesn't.

My goal is to support a distributed, extensible classification network -- 
which I think is the only way that any kind of standards can emerge. It's 
very much along the lines of what PeterV has outlined here (he and I have 
talked about it before), but I'm focusing on the software itself, and a 
data model so others can extend the software.

I'm explaining all this not just to bring it to your attention, but also to 
ask everyone what controlled vocabulary formats they like, and for 
conforming examples -- facetmaps are based on CVs. Any good but 
little-known resource classification method would be great too. A few have 
been mentioned on this list, but I'm looking for a broad sample of 
thesaurus markup (e.g. 
http://www.icpsr.umich.edu:8080/THESAURUS/thesaurus.dtd ), topic maps (e.g. 
XTM, XFML), resource descriptors (e.g. RDF), and so on.

If anyone knows, this list does. I'm a first-time caller, long-time listener,

Travis Wilson
Development
http://facetmap.com