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

From: John J. Deneen (jjdeneen@netzero.net)
Date: Wed Oct 03 2001 - 12:57:30 PDT

  • Next message: Eric Armstrong: "[unrev-II] Recommendations??"

    I highly recommend reading the following article ("Earlier Hijackings
    Offered Signals That Were Missed") in context with R&D experiences with
    a software program called "Hyper-Object Substrate" that provides
    evidence that some of the problems with formalization were solved or
    mitigated with their approach to incremental formalization:

       * premature structuring — solved by not forcing users to initially
         enter formal information;
       * cognitive overhead — mitigated due to lower overhead for initial
         entry and demand-driven effort required for later formalization;
         and
       * tacit knowledge — mitigated by suggestion mechanisms which promote
         conscious understanding of previously tacit knowledge.

    "Earlier Hijackings Offered Signals That Were Missed"
    October 2, 2001, The New York Times
    < http://www.nytimes.com/2001/10/03/national/03WARN.html >

    From a lessons-learned and improvement needs/benefit perspective, "[T]he
    failure to heed these signs is "an indication of failure to put the
    pieces together," said Gerald B. Kauvar, who was the staff director of
    the commission headed by Vice President Al Gore on aviation security and
    safety after the crash of T.W.A. Flight 800 off Long Island in July
    1996.

    The authorities appeared to draw no lessons from the two attacks in
    1994. But one of them, in hindsight, had striking similarities to those
    of Sept. 11. ...

     ... But extrapolation may be a key lesson of Sept. 11. According to one
    executive at an aviation trade organization, "now we need to do that in
    every element of our life; that's how we fight this new war."

    Insightful analysis of these problems for SDS, OHS/DKR, Topic Map, etc
    designers, including getting explicit formalization (e.g., from aviation
    security and safety experts).
    F. Shipman and R. McCall, "Supporting Incremental Formalization with the
    Hyper-Object Substrate", ACM Transactions on Information Systems, 17, 2
    (April 1999), pp. 199-227. <
    http://www.csdl.tamu.edu/~shipman/tois-hos.pdf >

    F. Shipman, C. Marshall, M. LeMere, "Beyond Location: Hypertext
    Workspaces and Non-Linear Views", ACM Hypertext '99 Proceedings, pp.
    121-130. (won the Douglas Engelbart best paper award) <
    http://www.csdl.tamu.edu/~shipman/ht99.pdf >

     ... "The situation is ironic. On the one hand, users — and software
    designers — want proactive, high-functionality systems to support their
    daily work. On the other hand, users resist the formalization of
    information that is required for such higher-level functionality.

    At first glance it might seem that we are facing an intractable dilemma:
    users demand higher levels of functionality yet are unwilling to pay the
    inevitable price of increased formalization efforts. But the appearance
    of dilemma here is illusory. It derives from the mistaken notions that
    formalization is an all-or-nothing and all-at-once proposition. We argue
    here that an incremental approach to formalization is both acceptable to
    users and effective in getting far greater amount of formalized
    information into systems.

    With this approach formalization need not be done at the time of
    information input and exclusively by the user; it can be done during
    system use and with substantial support from the computer.

    In the following section we explain the approach we call incremental
    formalization. We then propose a software architecture for supporting
    this approach — an architecture instantiated in a prototype we call the
    Hyper-Object Substrate (HOS). Following this, we describe experiences in
    using HOS for several types of applications that benefit from
    incremental formalization. We conclude by summarizing the benefits of
    HOS and by identifying issues that arose from our experience with the
    system."



    This archive was generated by hypermail 2.0.0 : Wed Oct 03 2001 - 12:47:25 PDT