Method and apparatus for human -assisted adaptive outbound telephone call classification based on pattern recognition

• US 20040037397A1
• Filed: 08/20/2002
• Published: 02/26/2004
• Est. Priority Date: 08/20/2002
• Status: Active Grant

• Current Assignee: Avaya Incorporated
• Original Assignee: Avaya Technology LLC (Avaya Incorporated)
• Inventors: Roby, Paul Vinton
• Granted Patent: US 7,190,768 B2
• Time in Patent Office: Days
• US Class Current: 379/88.01
• CPC Class Codes: H04M 2201/40 using speech recognition
• CPC Class Codes: H04M 2203/2027 Live party detection
• CPC Class Codes: H04M 3/5158 in combination with automated outdialling systems devices for signalling identity of wanted subscriber H04M1/26

Abstract

A system for classifying outbound telephone calls is provided. According to the invention, voice messages received as a result of placing outbound telephone calls can be classified. In particular, autovoice messages, including network intercept messages, can be detected even in the absence of a special information tone. Furthermore, the present invention allows for the initial classification of voice messages to be performed by human operators, with subsequent recognition of classified messages by automated systems. Appropriate responses to recognized voice messages can be performed in substantially real time with dialer operations, or during subsequent dialer operations.

Claims

1. A method for classifying outbound telephone calls, comprising:

placing a telephone call;

detecting an autovoice;

extracting at least first voice pattern data from said detected autovoice;

comparing said at least first extracted voice pattern data to at least first known voice pattern data;

in response to a match between said at least first extracted voice pattern data and said at least first known voice pattern data, performing a first predetermined action; and

in response to a failure to match said at least first extracted voice pattern data to said at least first known voice pattern data, storing said first extracted voice pattern data.

2. The method of claim 1, further comprising retrieving said at least first known voice pattern data from a first database, wherein said step of storing said at least first extracted voice pattern data comprises storing said extracted voice pattern data in at least one of said first database and a second database.

3. The method of claim 1, further comprising:

comparing said stored at least first extracted voice pattern data to at least second extracted voice pattern data.

4. The method of claim 3, further comprising:

matching said stored at least first extracted voice pattern data to at least second extracted voice pattern data; and

scripting at least a second predetermined action.

5. The method of claim 4, wherein said step of matching said stored at least first extracted voice pattern data to said at least second extracted voice pattern data comprises matching at least a first portion of said at least first extracted voice pattern data to at least a first portion of said at least second extracted voice pattern data.

6. The method of claim 4, wherein said step of scripting is performed by a human operator.

7. The method of claim 4, wherein said at least second extracted voice pattern data is retrieved from a database.

8. The method of claim 4, wherein said matched voice pattern data and said at least a second predetermined action are stored in a first database.

9. The method of claim 1, wherein said step of storing said extracted voice pattern data further comprises storing a number called and a sound recording corresponding to said extracted voice pattern data.

10. A computer readable storage medium containing programming code for performing the steps of claim 1.

11. A method for classifying outbound autovoice call terminations, comprising:

receiving at least first autovoice pattern data;

comparing said at least first autovoice pattern data to at least second autovoice pattern data;

in response to a match between at least a portion of said first autovoice pattern data and at least a portion of said second autovoice pattern data, scripting a response to said matching portions of said autovoice pattern data; and

storing said matching portions of said autovoice pattern data and said scripted response.

12. The method of claim 11, further comprising;

permitting access to said stored autovoice pattern data and said scripted responses by at least a first party performing outbound dialing operations.

13. The method of claim 11, wherein said at least first autovoice pattern data is received from at least a first party performing outbound dialing operations

14. An apparatus for classifying outbound telephone calls, comprising:

at least a first computer storage medium;

at least a first computer processor in communication with said at least a first computer storage medium; and

at least a first communication channel interface in communication with said at least a first computer processor, wherein first extracted voice pattern data from an autovoice received at said communication channel interface is compared to second voice pattern data stored on said computer storage medium, wherein a first predetermined response is performed in response to a match between said first extracted voice pattern data and said second voice pattern data, and wherein said first extracted voice pattern data is stored on said computer storage medium in response to a failure to match said first extracted voice pattern data to said second voice pattern data.

15. The apparatus of claim 14, further comprising a second computer storage medium, wherein said first extracted voice pattern data is stored on said second computer storage medium in response to a failure to match said first extracted voice pattern data to said second voice pattern data.

16. The apparatus of claim 14, further comprising:

a user output;

a user input, wherein a human operator is provided with said first extracted voice pattern data and third voice pattern data by said user output, and wherein said operator enters a predetermined response to a match between said third voice pattern data and a received extracted voice pattern data with said user input.

17. The apparatus of claim 14, wherein said first communication channel interface is interconnected to the public switched telephony network.

18. An outgoing call classification apparatus, comprising:

means for storing first voice pattern data;

means for receiving second voice pattern data;

means for comparing said first voice pattern data to said second voice pattern data; and

means for storing said second voice pattern data in response to second voice pattern data that does not match said first voice pattern data.

19. The apparatus of claim 18, further comprising:

means for comparing said second voice pattern data to third voice pattern data;

means for providing at least one of said second voice pattern data and said third voice pattern data to a human operator if said means for comparing matches said second voice pattern data to said third voice pattern data;

means for associating a script prepared by said human operator with at least a portion of at least one of said second voice pattern data and said third voice pattern data; and

means for providing said script and associated voice pattern data to said means for storing.

20. A computer readable storage medium carrying instructions for performing a method, the method comprising:

comparing first voice pattern data to stored second voice pattern data;

in response to detecting a match between said first voice pattern data and said second voice pattern data, performing instructions associated with said second voice pattern data; and

in response to failing to detect a match between said first voice pattern data and said second voice pattern data, storing said first voice pattern data.

21. The computer readable storage medium of claim 20, the method further comprising:

comparing said stored first voice pattern data to stored third voice pattern data; and

in response to detecting a match between said stored first voice pattern data and said third stored voice pattern data, providing at least one of at least a portion of said first voice pattern data and at least a portion of said third stored voice pattern data to a scripting agent, whereby instructions for responding to a detected a match between compared voice pattern data are prepared.

22. The computer readable storage medium of claim 21, wherein said scripting agent comprises a human operator.

23. The computer readable storage medium of claim 21, the method further comprising:

storing instructions prepared by said scripting agent, wherein said instructions are performed in response to detecting a match between fourth voice pattern data and said third stored voice pattern data.

24. The computer readable storage medium of claim 20, wherein said first voice pattern data is received from a computer storage medium.

25. The computer readable storage medium of claim 20, wherein said first voice pattern data is received from a switched circuit telephony network.

26. The computer readable storage medium of claims 20, wherein said stored third voice pattern data comprises an operator intercept message.

27. The computer readable storage medium of claim 20, the method further comprising:

prior to said step of comparing first voice pattern data to second voice pattern data, detecting first voice pattern comprising an autovoice.

28. The computer readable storage medium of claim 27, wherein said autovoice comprises at least one of an operator intercept message and a number specific voice response.

29. The computer readable storage medium of claim 20, wherein said first voice pattern data is received as part of a telephone call, and wherein said step of comparing first voice pattern data to stored second voice pattern data is performed before said telephone call is disconnected.

30. The computer readable storage medium of claim 20, the method further comprising:

storing said first voice pattern data, wherein said first voice pattern data is received as part of a telephone call, and wherein said step of comparing first voice pattern data to stored second voice pattern data is performed after said telephone call is disconnected.

31. The computer readable storage medium of claim 20, wherein said first and second voice pattern data is extracted from a voice message.