Browse Prior Art Database

Automatic Categorization Tool for Organizing Brainstorming Comments

IP.com Disclosure Number: IPCOM000122828D
Original Publication Date: 1998-Jan-01
Included in the Prior Art Database: 2005-Apr-04
Document File: 4 page(s) / 159K

Publishing Venue

IBM

Related People

Golovchinsky, G: AUTHOR [+2]

Abstract

Disclosed is a method and tool for automating the categorization process of ideas collected in TeamKit/2 and GroupSystems V electronic brainstorming sessions.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 36% of the total text.

Automatic Categorization Tool for Organizing Brainstorming Comments

      Disclosed is a method and tool for automating the
categorization process of ideas collected in TeamKit/2 and
GroupSystems V electronic brainstorming sessions.

      Novel Advance over the Prior Art: The Automated Categorization
Tool performs a cluster analysis to group brief brainstorming
comments into categories.  Currently available tools only search for
key terms that are defined by the user.  The automated tool groups
comments based  on patterns of terms that it discovers in the common
text.

      Technical Description: Team Kit/2 and Group Systems V are
software packages used to support electronic meetings using the Team
Focus technique.  The technique often begins with a brainstorming
process that generates many ideas in the form of typed comments.  The
next step in the technique is for the meeting attendees to group the
comments into categories based on the ideas expressed in the
comments.  Typically, the group reviews the comments and places them
into appropriate categories using a tool called the Idea Organizer
(IO).  This process can be very time consuming because the meeting is
essentially a non-electronic meeting

      The Automated Categorization Tool is designed to be run as an
intermediate step between idea generation and idea organization.  It
reads the files produced during an Electronic Brainstorming (EB)
session and creates a set of files on the facilitator's workstation.
Its output  can be incorporated directly into the IO tool, in the
same manner as the  results of an EB session run with keywords is
imported into an IO session.

      Text Pre-processing - In addition to the standard report file
produced by the EB tool, the categorization tool uses three auxiliary
files in its parsing process.  Default versions of these files are
provided with the software, but they can be customized with any
standard editor (e.g., E, edlin, EPM, XEDIT, etc) to tailor them to
specific requirements.

      Prefixes and suffixes.  One file contains a list of prefixes
and suffixes that are removed before comparing words.  This supports
a more intelligent comparison that is not sensitive to verb tense,
plural forms, etc.  An extensive set of suffixes is provided with the
tool.  Prefixes and suffixes are also removed from synonyms (see
below) to allow greater flexibility when matching terms among
comments.

      Synonyms.  Another file contains a list of synonyms.  Entries
in this file create the semantic links between words that express the
same or similar concepts but are not related orthographically.  Also,
this file can be used to store conjugations of irregular verbs that
are not handled by simple suffix removal (e.g., take, took), common
abbreviations (e.g., mgmt for management), and common misspellings.
These constructs all serve to improve the correlation between the
syntactic (the level of the cluster analysis)...