Browse Prior Art Database

Intelligent Rule-Building Based on System Learning of User Actions on Unread Mail

IP.com Disclosure Number: IPCOM000118543D
Original Publication Date: 1997-Mar-01
Included in the Prior Art Database: 2005-Apr-01
Document File: 2 page(s) / 38K

Publishing Venue

IBM

Related People

Murray, JL: AUTHOR [+3]

Abstract

Disclosed is a methodology to allow intelligent rule-building based on system learning of user habits regarding the disposition of unread mail.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 79% of the total text.

Intelligent Rule-Building Based on System Learning of User Actions
on Unread Mail

      Disclosed is a methodology to allow intelligent rule-building
based on system learning of user habits regarding the disposition of
unread mail.

      Intelligent agent applications typically allow users to define
rules that allow systems to handle repetitive, predictable tasks
automatically.  The user specifies the conditions that must be met to
trigger the rule and the desired actions related to those conditions.

      The methodology disclosed here takes this processing one step
further by monitoring user behavior in handling mail in an attempt to
find patterns where the system could suggest a rule to the user.  The
challenge in this work, in order to make the suggested rules useful,
is to recognize patterns of behavior that have very few, if any,
exceptions and where the user acted on the mail item given only a
limited set of information, such as sender and subject line
information.  This can be  accomplished by focusing on the actions
users take on their unread mail.  By focusing on the unread mail, the
need to account for user decisions regarding action based on content
in the body of the note is  eliminated.  This, in turn, reduces the
threshold that the system must  monitor for to one, since the user
was able to make a decision regarding  the action based on a very
limited set of information (i.e., sender and  subject information).
This would indicate a ver...