Re: [unrev-II] Semantic Community Web Portal (Formality Harmful)

From: Eugene Eric Kim (
Date: Mon Sep 17 2001 - 00:36:40 PDT

  • Next message: Eugene Eric Kim: "Re: [unrev-II] Fwd: [xml-dev] Pakistan has an English-language weblog"

    On Sun, 16 Sep 2001, Eric Armstrong wrote:

    > I think the answer is to create tools that let us think the way we
    > think, and then
    > go back an analyze the way we thought and restructure it to represent
    > the way
    > we would have like to have been thinking. That is, in essence, what we
    > do when
    > we write, design, create, etc. We put up an initial draft, and then
    > revise it into
    > something useful.
    > See the post on Eugene's work, next.

    Eric did a nice job of summarizing my experiments with dialog maps. There
    are a bunch of threads on this list to which I would like to respond, but
    I'll start here and work backwards.

    A few words on the methodology I've been experimenting with. It's very

    1. People discuss things in an unstructured fashion -- e-mail, newsgroups,
       web forums, whatever.

    2. One or more moderators build a dialog map of the discussion, with
       granular links to the original discourse.

    3. Partipants frequently check the dialog map, and have the option of
       commenting on it in their unstructured forums or even in the dialog map

    Some comments:

    * The resulting dialog map not only serves as a way for guiding future
      discussion, but also as an excellent snapshot of the existing
      discussion. Try looking at the unrev-ii archives sometimes. Like most
      mail archives, it's essentially useless unless you have several months
      of free time to wade through every single e-mail. The dialog map offers
      a readable summary of the discussion with links to the original
      discourse, so you can quickly identify threads you should be following.

    * IBIS can be both incredibly easy and incredibly difficult to use. A
      common phenomenon in facilitated IBIS discussions is that an initial
      question quickly spurs a number of ideas. However, as the discussion
      progresses, the grammar's constraints make it increasingly difficult to
      build and maintain the map.

      This is to be expected. If creating IBIS dialog maps were trivial, then
      it wouldn't really help improve discourse. The value of IBIS is that it
      forces you to consider the underlying questions, most of which are
      rarely obvious. Doing dialog maps well requires a skilled facilitator.

      My methodology allows you to edit the dialog map directly when it's
      easy, and use the unstructured mediums when it's not. Either way, the
      content is recorded in the dialog map, thanks to the diligence of the
      facilitators. This is really an extension of Jeff Conklin's methodology
      of IBIS as a facilitation technique, applied towards asynchronous

    * For this methodology to work effectively, it requires frequent
      validation by the participants. For this to happen, we need better

    As Eric noted, I've been using QuestMap primarily. I've also experimented
    extensively with Nexist, and I wrote some conversion tools in Perl, which
    allowed me to tinker with an underlying data model for IBIS. Eric also
    posted some design notes several months ago, which I found very useful.

    Based on my experiences with IBIS, the existing tools, and my methodology
    experiments, I've started to put together a design for a collaboration
    tool to support this structured/unstructured form of dialog. My first
    step will be to build a set of Java classes that implements an underlying
    IBIS data model. I plan on using these classes to build my own dialog
    mapping tool, but I hope that others will find it useful as well. For
    instance, I have some thoughts on how to integrate Alex's TouchGraph with
    these classes, but he will probably have a clearer idea on how this could
    be done, and he'd be free to use my library to demonstrate this. The more
    people who play with the classes, the better the APIs will become.

    I also plan on designing these classes in such a way that tools such as
    NODAL or Eliot Kimber's version-control grove engine could easily be
    plugged in. Finally, I plan on constructing a link database that could
    become a candidate link database for the OHS.

    I'll post more on my efforts as time permits.


    +=== Eugene Eric Kim ===== ===== ===+
    |       "Writer's block is a fancy term made up by whiners so they        |
    +=====  can have an excuse to drink alcohol."  --Steve Martin  ===========+

    ------------------------ Yahoo! Groups Sponsor ---------------------~--> Do you need to encrypt all your online transactions? Secure corporate intranets? Authenticate your Web sites? Whatever security your site needs, you'll find the perfect solution here! ---------------------------------------------------------------------~->

    Community email addresses: Post message: Subscribe: Unsubscribe: List owner:

    Shortcut URL to this page:

    Your use of Yahoo! Groups is subject to

    This archive was generated by hypermail 2.0.0 : Mon Sep 17 2001 - 03:53:35 PDT