Autonomous learning and reasoning agent

• US 5,852,814 A
• Filed: 12/10/1996
• Issued: 12/22/1998
• Est. Priority Date: 03/04/1991
• Status: Expired due to Fees

• Current Assignee: eGain Communications Corporation
• Original Assignee: Inference Corporation
• Inventors: Allen, Bradley P.
• Primary Examiner(s): Downs, Robert W.
• Application Number:
  US08/763,005
• Time in Patent Office: 742 Days
• Field of Search: 395/13, 706/13, 706/54
• US Class Current: 706/13
• CPC Class Codes: G06N 3/004 Artificial life, i.e. computing arrangements simulating life
• CPC Class Codes: G06N 3/126 Evolutionary algorithms, e.g. genetic algorithms or genetic programming
• CPC Class Codes: G06N 5/04 Inference or reasoning models

Abstract

A software agent which performs autonomous learning in a real-world environment, implemented in a case-based reasoning system and coupled to a sensor for gathering information from, and to an effector for manipulating, its environment. A case base which is tuned in response to an evaluation of how well the agent is operating in that environment. A memory of cases, the contents of that memory being determined by a genetic technique, including producing new cases which may never have been encountered in the environment, evaluating cases in response to a history of previous matches and in response to an external stimulus, and selecting a limited set of cases which provides a preferred model of the environment. A selection technique based on multiple measures, such as match quality, case accuracy, or case utility, which may induce experimentation by the autonomous agent, such as by employing a random or pseudorandom effect in selecting cases, and which may also distinguish between those actions which solve problems and those actions which gather further information so as to better solve problems. Multiple autonomous agents may form a collective entity and may cooperate to select an action to be performed by that collective entity.

Claims

1. In a case-based reasoning system, a method of altering a case base having a plurality of exemplar cases;

said method comprising

generating a case which has not been encountered before;

retrieving a set of matching cases from said case base;

evaluating said case in response to (a) said set of matching cases, or (b) said set of matching cases and a reinforcement received from an environment;

modifying weights of said set of matching cases to reflect the accuracy or inaccuracy of the evaluation; and

selecting a limited set of cases in which the evaluation was inaccurate for addition into said case base which provides a preferred model of said environment.