Tools for Dialogue and Deliberation on Wicked Problems

Subject-centric computing

For several years now, I've been giving talks on subject-centric computing with my talks on "federation". I've been amazed at how many people require that I define "federation". Now, it would seem that "Subject-Centric Computing" has come to have its day in the sun--likely through no efforts of mine, but still very satisfying. One of my earliest critics is now one of the loudest champions. Read Steve Pepper's slides to see what I mean. The paper is a nice overview of the issues related to subject-centric (as compared to document-centric) computing, but glosses over what I consider the primary issue: subject identity.

Key take-home point: migrate from document-centric to subject-centric and you remove an impressive amount of sting associated with infoglut. My view: we miss something when we talk about documents; we are really talking about subjects entrained in those documents. I could say "It's the Subject, Stupid!", but I won't.

Comment by David Price on April 29, 2008 at 0:01: Jack,

Thanks for pointing to the presentation—and can't wait to have a detailed discussion with you about this.

David

Comment by Jack Park on April 29, 2008 at 0:36: Would be nice to have a "real" (in the KMi sense) place to have such a discussion...

Comment by Jack Park on April 29, 2008 at 0:36: Hah! What about DebateGraph?

Comment by Mark Szpakowski on May 7, 2008 at 21:43: "It's the Subject, Stupid!" is kind of doubly self-referential.

From a kind of syntactical point of view the compound document folks of a couple of decades ago were on the same track. A document is made out of multiply-reference-able Addressable Information Objects (AIRS) - the pieces have permalinks. You're talking about document as embedding context for subject representations. And the framework for resolving subjects.

Something like this could be used to see if apples and apples are being compared when evaluating pro- and con- co-premises (thinking of the discussion thread on Collaborative Climate simulation models (wish I could link it into here more simply)).

Comment by Jack Park on May 7, 2008 at 22:43: But, I didn't say that!

In writing a paper for the upcoming TMRA workshop in Leipzig, I described the process this way: we bind subjects to documents by telling stories. We later harvest those subjects into indexes and maps. We later rebind those subjects into other documents by telling more stories about them. Documents are vehicles, or embedding context. Agree.

Evaluating pro and con arguments, their premises, etc, is an exercise in subject identification partially facilitated by the addressable documents that carry them. I'm not sure we have anything like a best practices for subject identification. Simple URIs for documents do more to locate the document but appear to say little, if anything, about the subject(s) entrained in the document.

Please talk more about using documents as frameworks for resolving subjects. Thanks!