Method of providing speech recognition for IVR and voice mail systems

• US 7,062,019 B2
• Filed: 07/27/2001
• Issued: 06/13/2006
• Est. Priority Date: 07/27/2001
• Status: Expired due to Term

• Current Assignee: Avaya Incorporated
• Original Assignee: Avaya Technology Corporation Miami Lakes FLA US
• Inventors: Krack, Mike
• Primary Examiner(s): Tsang, Fan
• Assistant Examiner(s): Gauthier, Gerald
• Application Number:
  US09/917,576
• Publication Number:
  US 20030021394A1
• Time in Patent Office: 1,782 Days
• Field of Search: 379/88.01, 379/88.1, 379/10.01, 379/88.18, 379/88.02, 379/88.03, 379/88.04, 379/74, 379/75, 379/76, 379/77
• US Class Current: 379/88.04
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 3/533 Voice mail systems

Abstract

The present invention is directed to a speech enabling gateway system that can provide a speech recognition capability to signal-based adjunct processors, such as DTMF-based adjunct processors. The gateway, in one configuration, conferences two or more channels together and performs speech recognition on data streams carried by one or both of the channels. Spoken commands are mapped to appropriate signals for the signal-based adjunct processor.

Claims

1. An interactive voice response system, comprising:

an adjunct processor that outputs an output data stream to a user; and

a speech gateway enabling system comprising;

a speech recognition engine operable to identify words in an input voice stream received from the user on a first communication path extending between the user and the speech gateway enabling system and

a speech gateway controller operable (a) to transfer at least a portion of the input voice streams, received from the users, from the first communication path to a second communication path extending between the speech gateway enabling system to the adjunct processor and (b) to transfer the at least a portion of the input voice stream from the first communication path to the speech recognition engine for processing.

2. The interactive voice response system of claim 1, wherein the speech gateway enabling system comprises a speech digitizer that converts the input voice stream from analog to digital form and the first and second communication paths are tromboned together.

3. The interactive voice response system of claim 1, wherein the first and second communication paths are configured by a switching system and the speech gateway controller is further operable to generate and transmit a command signal to the adjunct processor based on words identified by the speech recognition engine.

4. The interactive voice response system of claim 3, wherein the command signal is a DTMF code and the speech enabling gateway system comprises grammar correlating a plurality of words with a corresponding plurality of DTMF codes in the command set of the adjunct processor.

5. The interactive voice response system of claim 3, wherein the switching system comprises a plurality of communication ports and the first communication path extends between first and second communication ports of the switching system and the second communication path extends between different third and fourth communication ports of the switching system.

6. The interactive voice response system of claim 1, wherein the speech gateway controller performs operation (b) in a first operational mode and wherein the speech enabling gateway controller is also operable (c) to transfer at least a portion of the output data stream from the second communication path to the first communication path and (d), in a second operational mode, to transfer the at least a portion of the output data stream from the second communication path to the speech recognition engine for processing.

25. The interactive voice response system of claim 6, wherein the first operational mode is performed in response to a command signal from the user and the second operational mode is performed in response to a command signal from the adjunct processor.

26. The interactive voice response system of claim 3, wherein the speech gateway controller is operable to place the user on hold while the command codes are transmitted to the adjunct processor.

7. A method, comprising:

(a) directing to a speech recognition engine at least a portion of an input voice stream received from a user on a first communication path extending between the user and a first adjunct processor;

(b) detecting, with the speech recognition engine, at least some of the words in the at least a portion of the input voice stream;

(c) transferring the input voice stream from the first communication path to a second communication path extending between the first adjunct processor and a second adjunct processor;

(d) comparing at least some of the detected words with a grammar, the grammar correlating a plurality of words with a corresponding plurality of command codes, to identify corresponding command codes for each of the at least some of the detected words; and

(e) transmitting a command signal corresponding to at least one identified command code to the second adjunct processor on the second communication path.

8. The method of claim 7, wherein the directing and transferring steps occur at least substantially simultaneously.

9. The method of claim 7, wherein the grammar further includes at least one switch symbol for at least one of enabling and disabling the directing step (a).

10. The method of claim 7, further comprising:

(f) converting the input voice stream from an analog form to a digital form.

11. The method of claim 7, further comprising:

(f) configuring the first communication path for a first communication session initiated by the user with the first adjunct processor; and

(g) thereafter configuring the second communication path for a second communication session, initiated by the first adjunct processor, between the first and second adjunct processors.

12. The method of claim 7, further comprising:

(f) transferring an output data stream from the second communication path to the first communication path, the output data stream being received from the second adjunct processor.

13. The method of claim 12, wherein transferring steps(c) and (f) occur at least substantially simultaneously.

14. The method of claim 7, further comprising:

(f) muting the first communication path when the transmitting step (e) is performed.

15. The method of claim 7, further comprising:

(f) determining if the output from the speech recognition engine includes a switch symbol; and

(g) when the output includes a switch symbol, at least one of enabling or disabling the directing step (a).

16. The method of claim 7, further comprising:

(f) determining if one of the first and second communication paths has been disconnected; and

(g) when one of the first and second communication paths has been disconnected, disconnecting the other of the first and second communication paths.

27. The method of claim 7, wherein the plurality of command codes are DTMF codes and the command signal is a DTMF signal.

28. The method of claim 12, further comprising:

(g) directing to the speech recognition engine at least a portion of the output data stream received from the second adjunct processor on the second communication path extending between the first and second adjunct processors;

(h) determining when the output data stream includes a switch symbol;

(i) when the output data stream includes a switch symbol, directing step (g) is performed and, when the input voice stream includes a switch symbol, directing step (a) is performed.

40. The method of claim 28, wherein the directing steps (a) and (g) occur at different times and the transferring steps (c) and (f) occur at least substantially simultaneously.

17. A system, comprising:

first and second adjunct processors;

a speech recognition engine that detects at least some words in an input voice stream received from a user on a first communication path extending between the user and the first adjunct processor;

comparing means for comparing at least some of the detected words with a grammar, the grammar correlating a plurality of words with a corresponding plurality of DTMF codes, to identify corresponding DTMF codes for each of the at least some of the detected words;

directing means for directing to the speech recognition engine at least a portion of the input voice stream received from the user on the first communication path;

transferring means for transferring the at least a portion of the input voice stream received from the first communication path to a second communication path extending between the first adjunct processor and the second adjunct processor; and

transmitting means for transmitting a DTMF signal corresponding to at least one identified DTMF code on the second communication path.

18. The system of claim 17, wherein the grammar further includes at least one switch symbol for at least one of enabling and disabling the directing means.

19. The system of claim 17, further comprising:

converting means for converting the input voice stream from an analog form to a digital form.

20. The system of claim 17, further comprising:

configuring means for configuring the first communication path for a first communication session initiated by the user with the first adjunct processor and thereafter configuring the second communication path for a second communication session, initiated by the first adjunct processor, between the first and second adjunct processors.

21. The system of claim 17, further comprising:

transferring means for transferring an output data stream from the second communication path to the first communication path.

22. The system of claim 17, further comprising:

muting means for muting the first communication path when the transmitting means transmits DTMF signals.

23. The system of claim 17, further comprising:

determining means for determining if the output from the speech recognition engine includes a switch symbol; and

when the output includes a switch symbol, means for at least one of enabling or disabling the directing means.

24. The system of claim 17, further comprising:

means for determining if one of the first and second communication paths has been disconnected; and

when one of the first and second communication paths has been disconnected, means for disconnecting the other of the first and second communication paths.

41. The system of claim 17, wherein the directing and transferring operations occur simultaneously.

29. A method, comprising:

(a) directing to a speech recognition engine at least a portion of an output data stream received from a second adjunct processor on a second communication path extending between the second adjunct processor and a first adjunct processor;

(b) detecting, with the speech recognition engine, at least some of the words in the at least a portion of the output data stream received from the second adjunct processor;

(c) transferring the at least a portion of an output data stream received from the second adjunct processor to a first communication path extending between the user and the first adjunct processor;

(d) comparing at least some of the detected words with at least one command signal; and

(e) when the output data stream includes a command signal, terminating the directing step.

30. The method of claim 29, further comprising:

(f) directing to the speech recognition engine at least a portion of an input voice stream received from a user on the first communication path extending;

(g) detecting, with the speech recognition engine, at least some of the words in the at least a portion of the input voice stream;

(h) transferring the input voice stream to the second communication path;

(i) comparing at least some of the detected words with a grammar, the grammar correlating a plurality of words with a corresponding plurality of command codes, to identify corresponding command codes for each of the at least some of the detected words; and

(j) transmitting a command signal corresponding to at least one identified command code on the second communication path.

31. The method of claim 29, wherein the directing and transferring steps occur at least substantially simultaneously.

32. The method of claim 30, wherein the grammar further includes at least one switch symbol for at least one of enabling and disabling the directing steps (a) and (f).

33. The method of claim 30, further comprising:

(k) converting the input voice stream from an analog form to a digital form.

34. The method of claim 30, further comprising:

(k) configuring the first communication path for a first communication session initiated by the user with the first adjunct processor; and

(l) thereafter configuring the second communication path for a second communication session, initiated by the first adjunct processor, between the first and second adjunct processors.

35. The method of claim 29, wherein transferring step (c) and the directing step (a) occur at least substantially simultaneously.

36. The method of claim 30, wherein the directing steps (a) and (f) occur at different times and the transferring steps (c) and (h) occur at least substantially simultaneously.

37. The method of claim 30, further comprising:

(f) muting the first communication path when the transmitting step (j) is performed.

38. The method of claim 30, further comprising:

(k) determining when the at least a portion of the input voice stream includes a switch symbol; and

(l) when the at least a portion of the input voice stream includes a switch symbol, at least one of enabling or disabling the directing step (f).

39. The method of claim 30, further comprising:

(f) determining when one of the first and second communication paths has been disconnected; and

(g) when one of the first and second communication paths has been disconnected, disconnecting the other of the first and second communication paths.

42. The method of claim 30, wherein directing step (f) and transferring step (h) occur simultaneously.