Re: [unrev-II] problems with data retrieval by keyword

From: Eric Armstrong (eric.armstrong@sun.com)
Date: Thu Nov 08 2001 - 13:59:51 PST

  • Next message: Eric Armstrong: "[unrev-II] Graphics with hypertext!"

    Alex Shapiro wrote:

    > Here is the link:
    >
    > http://nooface.com/articles/01/11/06/1712235.shtml
    >
    > "...as the amount of searchable data soars, the retrieval process can
    > become inefficient if the user's vocabulary does not match the index
    > vocabulary used by the search process, or the user can not easily
    > constrain a search path based on context
    > ... This paper makes the case that the context problem can
    > be overcome through data visualization, which provides an additional
    > dimension for the user to specify their desired context...."

    It certainly sounds pretty good. But since you evaluate the paper as not
    that
    good, let me get your thoughts on a few questions:
      1. vocabulary mismatch and constraint based on context
          -- is that one problem, or two?

      2. If two, I take that no solution is being proposed for the first
    problem?

      3. In either case, I take it that ability to constrain the search wrt
    (with respect to)
          context means, for example, looking for looking for "shoes" in the
    context
          of "marching" rather than "dance", for example -- and further
    distinguishing
          context as pertaining to "marching band" rather than "forced night
    march".

       4. You have probably answered this question at length already, but
    now that
           can relate it to a specific problem:
              a) How do data visualization tools help to constrain a search?

              b) Assuming that they do make a difference, aren't they in
    fact doing
                  the kind of automated indexing that would also make
    constraint-
                  restrictions possible?
              c) Is the difference between "data visualization" and
    "automated indexing",
                  then, simply a matter of dynamic, client side indexing
    with visualization
                  tools, vs. relatively static server-side work with
    automated indexing
                  tools? (Graphic presentations aside, of course -- I'm
    assuming that
                  any visual representation is irrelevant for the purposes
    of constraining
                  a search. Or is there some way in which visual
    representation is
                  relevant?)

    Community email addresses:
      Post message: unrev-II@onelist.com
      Subscribe: unrev-II-subscribe@onelist.com
      Unsubscribe: unrev-II-unsubscribe@onelist.com
      List owner: unrev-II-owner@onelist.com

    Shortcut URL to this page:
      http://www.onelist.com/community/unrev-II

    Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/



    This archive was generated by hypermail 2.0.0 : Thu Nov 08 2001 - 13:47:23 PST