Site Search:

Contact Us


(Register Here)
DOD TechMatch Sponsors
Patent View
 Back to Listing
Patent 7925605 Click For Printable Version of This Patent
Patent Information:  USPTO Site Listing

Patent Number: 7925605
Case ID: 0
Patent Title: Evolutionary expert systems and methods using meta-rules matching
Status: ACTIVE
Status Date: 11/30/2011 10:09:21 AM
Issue Date: 4/12/2011
Filed Date: 9/13/2007
Serial #: 1/854,638
Assignee Name: The United States of America as represented by the Secretary of the Navy (Washington, DC)
Inventor(s): Rubin, Stuart H.
Lab Information:  View Lab Profile

Lab Name: SPAWAR Systems Center, Pacific
Location: 53560 Hull Street
San Diego, CA 92152-5001
Contact: Contact Lab About This Patent

I. Field

This disclosure relates to expert systems and other information processing apparatuses and methods.

II. Background

An expert system can be described as a program or system that contains some subject-specific knowledge and emulates human reasoning. Typically, expert systems include a user interface, a knowledge base, an inference engine and a data base. The user interface provides a portal for a user to interact with the expert system; the data base feeds data to the expert system; the knowledge base contains the rules that control the reasoning process; and the inference engine itself processes the rules inside the knowledge base to emulate reasoning.

An example of an expert system might be a banking system configured to process loan applications. After an applicant files a loan application, the data from the loan application can be fed to the knowledge base, where rules embedded within the expert system could determi . . . . More
In various embodiments, evolutionary expert systems and methods are disclosed. For example, a method for evolving a rule base of an expert system includes creating a set of meta-rules from a set of first rules associated with the expert system, creating a set of one or more generalized virtual rule candidates based on the set of first rules and the set of meta-rules, filtering the set of generalized virtual rule candidates to remove generalized virtual rule candidates that conflict with at least one rule of the set of first rules to form a set of virtual rules, and incorporating at least one virtual rule of the set of virtual rules into the set of first rules to evolve the first set of rules.
What is claimed is:

1. A method for evolving a rule base of an expert system on a computer, the method comprising: populating an application rule base stored in the computer with deterministic and semantically normalized application rules, wherein each application rule is of the form f.sub.0, f.sub.1, f.sub.2, . . . .fwdarw.p.sub.0, p.sub.1, p.sub.2, . . . ; while the computer is in a dream mode, automatically populating a meta-rule base with nondeterministic meta-rules of the form f.sub.0, f.sub.1, f.sub.2, . . . .fwdarw.p.sub.0, p.sub.1, p.sub.2, . . . f''.sub.0, f''.sub.1, f''.sub.2, . . . .fwdarw.p''.sub.0, p''.sub.1, p''.sub.2, . . . , wherein each meta-rule maps to two distinct application rules; in response to a query regarding a given context, searching the application rule base for an application rule that has a matching context; if no application rule is found with a matching context, deriving from a sequence of meta-rules a virtual rule with a matching conte . . . . More