Feature interaction resolution using fuzzy rules

• US 6,915,285 B2
• Filed: 05/21/2003
• Issued: 07/05/2005
• Est. Priority Date: 07/09/1999
• Status: Expired due to Term

• Current Assignee: Mitel Networks Corporation
• Original Assignee: Mitel Corporation (Microchip Technology Incorporated), University of Ottawa
• Inventors: Ahmer, Magdi, Mankovskii, Serge, Gray, Tom, Karmouch, Ahmed
• Primary Examiner(s): Hirl, Joseph P.
• Application Number:
  US10/442,571
• Publication Number:
  US 20030200187A1
• Time in Patent Office: 776 Days
• Field of Search: 706/49, 706/46, 706/47
• US Class Current: 706/49
• CPC Class Codes: G06N 5/048 Fuzzy inferencing
• CPC Class Codes: H04M 3/4217 Managing service interactions
• CPC Class Codes: H04Q 3/0041 involving techniques for avoiding interaction of call service features
• CPC Class Codes: Y10S 706/902 Application using ai with detail of the ai system

Abstract

A method of detecting feature interaction conflicts between agents in response to events in a multi-agent system comprises posting said events as tuples and event requests as anti-tuples from said agents in said multi-agent system, receiving said events and said event requests in a tuple space, generating responses to each of said agents for said events which match said event requests of said agents, wherein feature interaction conflict is detected when more than one of said agents are to exert control over an event.

Claims

1. A computer-implemented method of detecting feature interaction conflicts between agents in response to events in a multi-agent system, comprising:

posting said events as tuples and event requests as anti-tuples from said agents in said multi-agent system;

receiving said events and said event requests in a tuple space; and

generating responses to each of said agents for said events which match said event requests of said agents; and

detecting a feature interaction conflict when more than one of said agents are to exert control over an event.

2. The computer-implemented method of claim 1, wherein feature interaction conflict is also detected when an agent requesting an event refines permission for a suggested action by another agent exerting control over the event.

3. The computer-implemented method of claim 1, wherein said event requests are posted as one of either event notification by which said agents requesting said event requests are to be notified of an occurrence of certain events, or event handling by which said agents requesting said event requests are to exert control over associated events.

4. The computer-implemented method of claim 2, wherein said event requests are posted as one of either event notification by which said agents requesting said event requests are to be notified of an occurrence of certain events, or event handling by which said agents requesting said event requests are to exert control over associated events.

5. The computer-implemented method of claim 1, further comprising invoking an arbitration mechanism for resolving said feature interaction conflict over said events between said agents.

6. The computer-implemented method of claim 2, further comprising invoking an arbitration mechanism for resolving said feature interaction conflict over said events between said agents.

7. The computer-implemented method of claim 3, further comprising invoking an arbitration mechanism for resolving said feature interaction conflict over said events between said agents.

8. The computer-implemented method of claim 4, further comprising invoking an arbitration mechanism for resolving said feature interaction conflict over said events between said agents.

9. A computer-implemented event based model for detecting feature-interaction conflict in a multi-agent system comprising:

an Event Agent for posting events as tuples and event requests as anti-tuples from agents in said multi-agent system; and

a Blackboard having a tuple space for receiving said events and said event requests posted by said Event Agent and generating responses to each of said agents for said events which match said event requests of said agents; and

wherein the computer-implemented event based model is operable to detect a feature interaction conflict when more than one of said agents are to exert control over an event.

10. The computer-implemented event based model of claim 9, wherein feature interaction conflict is also detected when an agent requesting an event refuses permission for a suggested action by another agent exerting control over the event.

11. The computer-implemented event based model of claim 9, wherein said Event Agent includes means for posting said event requests as one of either event notification by which said agents requesting said event requests are to be notified of an occurrence of certain events, or event handling by which said agents requesting said event requests are to exert control over associated events.

12. The computer-implemented event based model of claim 10, wherein said Event Agent includes means for posting said event requests as one of either event notification by which said agents requesting said event requests are to be notified of an occurrence of certain events, or event handling by which said agents requesting said event requests are to exert control over associated events.

13. The computer-implemented event based model of claim 9, further comprising an Arbitrator for resolving conflicting control over said events between said agents.

14. The computer-implemented event based model of claim 10, further comprising an Arbitrator for resolving conflicting control over said events between said agents.

15. The computer-implemented event based model of claim 11, further comprising an Arbitrator for resolving conflicting control over said events between said agents.

16. The computer-implemented event based model of claim 12, further comprising an Arbitrator for resolving conflicting control over said events between said agents.

17. In a multi-agent system using a tuple space for communicating between a plurality of agents, a computer-implemented method of resolving feature interaction conflicts between said agents, the computer-implemented method comprising:

assigning fuzzy parameters of various degrees of truth for defining user preferences to attributes of rules associated with responses of said agents and to events of said agents;

detecting a feature interaction conflict for an event of a requesting agent posted on said tuple space which causes at least one conflicting response from at least one recipient agent;

opening a communications channel on said tuple space for posting said event with a degree of truth from said requesting agent and suggested responses with said degrees of truth from said at least one recipient agent; and

comparing said degree of truth and said degrees of truth of respectively said event and said suggested responses, and then selecting one of said suggested responses having the highest degree of truth as a response to said event.

18. The computer-implemented method of claim 17, wherein said at least one recipient agent provides at least one suggested response.

19. In a multi-agent system using a tuple space for communicating between a plurality of agents, a computer-implemented event based model for resolving feature interaction conflicts between said agents, the event based model comprising:

an Event Agent for posting events of said agents where said agents assign fuzzy parameters of various degrees of truth for defining user preferences to attributes of rules associated with responses of said agents and to said events of said agents;

a Blackboard having said tuple space for receiving said events posted by said Event Agent; and

an Arbitrator for detecting an feature interaction conflict to an event of a requesting agent posted on said Blackboard which causes at least one conflicting response from at least one recipient agent, and for opening a communications channel on said tuple space to post said event with a degree of truth from said requesting agent and suggested responses with said degrees of truth from said at least one recipient agent, and comparing said degree of truth and said degrees of truth of respectively said event and said suggested responses, and then selecting one of said suggested responses having the highest degree of truth as a response to said event.

20. The computer-implemented event based model of claim 19, wherein said at least one recipient agent provides at least one suggested response.

21. The computer-implemented event based model of claim 19, wherein the plurality of agents that are not participating in the feature interaction conflicts are not guaranteed to respond to the selected one of said suggested responses.