Browse Prior Art Database

A method for automating the categorization and storing of mail items in an electronic mail system Disclosure Number: IPCOM000031186D
Original Publication Date: 2004-Sep-16
Included in the Prior Art Database: 2004-Sep-16
Document File: 2 page(s) / 41K

Publishing Venue



This article documents the use of an artificial intelligence engine to drive personalization of a computer system, in this example an email system. The personalization engine allows for the automation of the personalization as the IE engine learns from the users behavior in interactions with the target system and uses this learning to personalize the views of the system presented to the user. This allows for changes in interests and therefore interaction behaviors with the system by the user over time.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

A method for automating the categorization and storing of mail items in an electronic mail system

Main Idea

Personalization of the user interface of various computer systems has become an almost daily task for many knowledge workers. The requirement to deal with general systems and to gain the most effective use of them to support the productivity of the business professional has driven the growth of this task. Many systems provide some form of personalization capability built in to the system, but these tend to be a static set of options and settings that have little or no intelligence behind them. One prime example of such a system is the electronic mail systems used in almost all businesses today.

With the volume of business related electronic mail increasing at a very high rate, more and more time is spent by professionals categorizing and storing their mail items so as to be able to retrieve related mail items over time as required. In addition an ever increasing volume of unwanted mail based advertizing (A.K.A. spam) is being recieved and while there is the possibility of using e-mail system agents (programmed functions) to delete this mail the originators are using various techniques to circumvent these fixed logic mail agents. More and more people are spending significant amounts of time during the day to manage their mail file.

This publication describes a solution which ties an artificial intelligence system to a computer system, in the example an electronic mail system) with some predefined logic and a starter set of simple rules to allow the system to automatically categorize, store and delete mail items. The system has two basic modes of operation, initial learning mode and operational mode. In initial learning mode the system monitors the electronic mail system and "learns" from the mail box owners actions on the incoming mail. In operational mode the system applies the users preferences as witnessed during the initial learn mode and continues to learn as the user takes different or new actions over time. A log of automated actions is kept for review by the user, should the user make any changes to the automated actions taken by the automated system, the system recognizes those changes and adjusts its knowledge base for that user as appropriate.

In the example the system (comprised of one or more computer software programs) resides either on the mail server (the central repository of multiple users electronic mail), or on a ma...