[Sigia-l] Information-centered Design

[Joe 10]

At 2:55 PM -0500 4/3/03, Matthew Rehkopf wrote:

Amazing thing about these threads, is that everyone seems to miss 
that almost every entry on the W3C home page has to do with 
"semantics". It's as if the IA community is percolating deep down 
this LIS/UCD silo and ignoring the fact that some pretty smart eggs 
are putting their heads to figuring out ways of making these gears 
mesh. Maybe it's some unwritten rule that I'll get blackballed from 
the upcoming IA egg toss for mentioning it, but RDF and related W3C 
initiatives holds the promise of what Matthew is asking. Now granted, 
promises have been broken before...

One of the problems I see is that we (some of us anyway, to some 
degree) are, as JGG put it, artisans, and creative types tend to dig 
doing it their own way and bristle at the thought of being hemmed in 
by descriptive standards.

Another barrier I see is that precise descriptions of information 
types (classes, or whatever you want to call them) opens the 
potential for automation of structure,  searching and retrieval 
through IA's evil palindrome, AI, and that scares some people.

I recall one of the first formal IA drinking events I went to was 
adjourning and the host said "Let's decide what we want to talk about 
at the next event. Let your mind go crazy - throw out any idea... 
except artificial intelligence". Sounded like fear of replacement to 
me. But like any technological advancement, AI won't displace [good] 
people, it will just allow them to perform higher functions.

A third barrier is in the "grouper" vs. "splitter" style difference. 
IAs tend to be "groupers". When you describe an asset to the 
granularity of a triple (as done in the RDF scheme of things) it 
ceases being information and becomes data, thus the domain of the 
"splitter". It's a rare head who can work in both domains, but this 
crowd can do it if any can. I think RDF may have promise as the basis 
for polymorphous information design.

[SNIP]
>
><crazy radical soapbox speech>
>Here's my GUT of Information Architecture, in my very non-scientific speech:
>The future web will not contain "sites" built around business goals and user
>needs. Instead, it will be structured around information nodes connected by
>relevance to other nodes. IAs will develop and maintain these nodes,
>concerned only with the relevance of content.

Precisely the RDF concept. Describing "stuff" to ridiculous levels of 
granularity and doing so in a way that allows sharing will make the 
Internet (perhaps not the "Web" mind you) a big information 
playground over and above the big info sewer the Web has become.

>
>How is this different? Let's take an example. Today, we build, for example,
>furniture sites for companies building or selling furniture; separate sites
>for each brand, on and on they go, too many to list here. We build them for
>businesses that have one thing in mind - to make money; big surprise.
>Businesses control this information; our current job is to make it easy for
>furniture buyers to find what they are looking for (User-centered design may
>only exist to make content easy to find on *my* site), and that activity
>makes us all proud. However, there are many others sites on related topics
>like: building furniture yourself, reviews of furniture, styles of
>furniture, etc, but this information is controlled and developed by others,
>again thinking of only their own business goals, and, thus, disconnected
>from the rest of furniture sites.
>
>In the future, there will not be these independent brand sites. There will
>be *one furniture site* where all the brands will post their products. Why?

Maybe not post their products in the current 'push' sense, but if 
described sufficiently then their products will be more easily and 
accurately aggregated into search results far more accurate than 
those derived from fuzzy, algorithmic responses we rely on today.

[SNIP]

>2. Lou Rosenfeld asked Amy Warner about connecting vocabularies. Here it
>starts, if it has not started already. Vocabs defined by relevance,
>universal to all, the beginning of one furniture site, one book site, one
>medicine site, one food site, one anything site...
>

That's a cornerstone of the RDF concept. One description can (in fact 
is expected to) leach descriptions already defined elsewhere. I won't 
go into it here, just point to the Web Ontology last call working 
draft at:
http://www.w3.org/TR/2003/WD-owl-guide-20030331/

As always, seeking reflection and debate.

/Joe

-- 

Joe Tennis
Information Design Honcho
Joe 10 Group
2430 5th Street
Studio L
Berkeley, CA 94710
510-649-1744
joe @ joe10.com
http://www.joe10.com?cpn=sig
User Centered Design, Strategy and Marketing
for Web, Wireless and Interactive Media

Get your E-Metrics in order:
http://metrics.joe10.com/