Quality driven expert system

• US 5,784,539 A
• Filed: 11/26/1996
• Issued: 07/21/1998
• Est. Priority Date: 11/26/1996
• Status: Expired due to Term

• Current Assignee: Cufer Asset Ltd LLC (Intellectual Ventures LLC)
• Original Assignee: Client-Server Networking Solutions, Inc.
• Inventors: Lenz, Frederick P.
• Primary Examiner(s): Hafiz, Tariq R.
• Assistant Examiner(s): Myhre, James W.
• Application Number:
  US08/753,484
• Time in Patent Office: 602 Days
• Field of Search: 395/50, 395/917, 395/919
• US Class Current: 706/45
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

A quality driven expert system maps answers input by a user in a user questionnaire into a set of qualities descriptive of a desired output entity. A library of output entities, such as network architectures is provided, each described by values for various ones of the qualities. The qualities derived by the expert system from the user'"'"'s answers are matched against the qualities of the various output entities in the library to determine which entities match the specified qualities. A score for each output entity is determined by an associative matching algorithm as a function of the degree of match between each entity and the specified qualities, along with an explanation of the degree of match between each specified quality and the qualities of the output entity. A selected number of the highest scoring entities are returned to the user to further inspection and comparison. The expert system architecture may be applied in various application domains, one useful implementation of which is for the selection of network architectures for computer networks.

Claims

1. A quality driven expert system, executable on a computer system including a processor, the system comprising:

a user questionnaire module that presents a plurality of questions to a user and receives a plurality of answers;

an expert system module that maps the answers into a set of qualities to determine a set of required qualities, each required quality having a first quality value derived from at least one answer;

a database of output entities, each output entity defined by selected ones of the qualities, each selected quality of an output entity having a second quality value; and

a matching algorithm module that receives the required qualities from the expert system and compares the first quality values of the required qualities to the second quality values of selected ones of the qualities of the output entities to determine a ranked set of output entities that satisfy the required qualities.

2. The expert system of claim 1, wherein:

the matching algorithm module outputs for each selected output entity a score as a function of whether the second quality values of the selected ones of the qualities satisfy the first quality values of the required qualities.

3. The expert system of claim 2, wherein:

each required quality has a quality operator; and

the score of an output entity is determined by whether for each required quality, the each second quality value of a corresponding selected quality of the output entity satisfies the quality operator and first quality value of the required quality.

4. The expert system of claim 2, wherein:

a at least one required quality has a quality weight derived from the user'"'"'s answers; and

the score of each output entity is a function of the weight of at least one required quality.

5. The expert system of claim 1, wherein the qualities are orthogonal.

6. The expert system of claim 1, wherein at least one quality is hierarchically defined by at least two other qualities.

7. The expert system of claim 1, wherein the expert system includes at least one rule including an antecedent defining a value of a fact, and a consequent defining a value to be assigned to a quality.

8. A computer mediated process of selecting a network architecture for a computer network, comprising:

presenting a plurality of questions to a user regarding the user'"'"'s requirements for the computer network;

receiving from the user answers to the questions;

deriving from the user'"'"'s answers a set of required qualities for the network architecture, each required quality having a requird quality value;

providing a plurality of network architectures, each network architecture defined by selected ones of the set of qualities, each of the selected ones of the qualities having a defined quality value; and

selecting at least one network architecture by comparing the required values of the required qualities to the defined values of the qualities of a network architecture.

9. The method of claim 8, further comprising:

providing a selection explanation of each selected network to the user, the selection explanation describing for each required quality the degree to which the selected network architecture satisifies the required quality.

10. The method of claim 8, further comprising:

wherein selecting at least one network architecture by comparing the required values of the required qualities to the defined values of the qualities of a network architecture, comprises;

selecting at least one network architecture as a function of a difference between the required values of the required qualities and defined values of the qualities of the network architecture.

11. The method of claim 10, further comprising:

determining for at least one required quality a weight of the required quality; and

wherein the difference between a required value of a required quality and a defined value is weighted by the weight of the required quality.

12. The method of claim 8 wherein the qualities are orthogonal.

13. The method of claim 8 wherein at least one quality is hierarchically defined by at least two other qualities.

14. A quality driven expert system for selecting a network architecture, executable on a computer system including a processor, the system comprising:

a database of network architectures, each defined by selected ones of a set of qualities, each selected quality of an network architecture having a defined quality value;

a user questionnaire module that presents a plurality of questions to a user and receives a plurality of answers;

an expert system module that receives the answers from the user questionnaire module and derives from the answers a set of required qualities, each required quality having a required quality value;

a matching algorithm module, coupled to the database, that receives the required qualities from the expert system and determines a score for at least one network architecture from the database by determining for each required quality, whether the defined quality value of a quality of the network architecture corresponding to the required quality satisfies the quality value of the required quality, matching algorithm module ranking the network architectures by their scores.

15. The expert system of claim 14, wherein the matching algorithm produces for each network architecture that is scored a selection explanation, the selection explanation indicating for each required quality the degree to which the required quality was satisified by a corresponding quality of the network architecture.