Technique for recognizing telephone numbers and other spoken information embedded in voice messages stored in a voice messaging system

• US 6,570,964 B1
• Filed: 04/16/1999
• Issued: 05/27/2003
• Est. Priority Date: 04/16/1999
• Status: Expired due to Term

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Enthoven, Dan, Murveit, Hy
• Primary Examiner(s): Hoosain, Allan
• Application Number:
  US09/293,660
• Time in Patent Office: 1,502 Days
• Field of Search: 379/67.1, 379/68, 379/69, 379/70, 379/71, 379/72, 379/76, 379/80, 379/85, 379/88.01, 379/88.02, 379/88.03, 379/88.07, 379/88.11, 379/88.12, 379/88.19, 379/88.2, 379/88.21, 379/88.27
• US Class Current: 379/67.1
• CPC Class Codes: G10L 15/193 Formal grammars, e.g. finite state automata, context free grammars or word networks
• CPC Class Codes: H04M 1/6505 storing speech in digital form
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 2203/301 Management of recordings
• CPC Class Codes: H04M 3/44 Additional connecting arrangements for providing access to frequently-wanted subscribers, e.g. abbreviated dialling at the subscriber's set H04M1/27; automatic redialling H04M3/424
• CPC Class Codes: H04M 3/53333 Message receiving aspects

Abstract

A technique for recognizing telephone numbers and other information embedded in voice messages stored in a telephone voice messaging system. A voice recognition system is coupled to the telephone voice messaging system. A voice message stored in the voice messaging system is transferred to the voice recognition system. The voice recognition system segments the voice message and then searches the segments for a predetermined speech reference model (grammar) which is expected to contain information of importance to the recipient of the message. In a preferred embodiment, the predetermined is a numeric grammar which specifies a sequence of numbers occurring in the voice message. In alternate embodiments, the grammar specifies a date, a time, an address, and so forth, and can specify more than one such type of information. The grammar can be modified or selected by the recipient of the voice message so that the voice recognition system searches for information of particular interest to the recipient. Once the predetermined grammar is identified, the voice recognition system outputs a portion of the stored voice message which includes the grammar. The output can be a display of the information contained in the grammar, such as a telephone number or an address. Alternately, the output can be an audible replay of the portion of the stored voice message which includes the grammar.

Claims

1. A method of recognizing information embedded in a stored voice message, wherein the method comprises steps of:

a. providing the stored voice message to a speech recognizer;

b. searching the stored voice message for all occurrences of a plurality of predetermined grammar types wherein each predetermined grammar type is one of a numeric grammar, an alpha grammar and an alphanumeric grammar, and further wherein each found occurrence corresponds to a predetermined grammar;

c. identifying a portion of the stored voice message including the predetermined grammar; and

d. outputting the portion of the stored voice message including each found predetermined grammar.

2. The method according to claim 1 wherein the step of outputting comprises a step of audibly replaying the portion of the stored voice message which includes the predetermined grammar.

3. The method according to claim 1 wherein the voice message is stored in a telephone voice mail system.

4. The method according to claim 3 wherein the step of searching is performed while the stored voice message is being recorded by a caller.

5. The method according to claim 1 wherein the predetermined grammar is selected by a recipient of the stored voice message.

6. The method according to claim 5 wherein the recipient of message selects the predetermined grammar after listening to the stored voice message.

7. The method according to claim 5 wherein the predetermined grammar is selected from a group consisting of:

a telephone number grammar, an address grammar, a date grammar, a person'"'"'s name grammar and a time of day grammar.

8. The method according to claim 1 further comprising a step of searching for at least one additional predetermined grammar.

9. A The method according to claim 8 wherein the predetermined grammar and the at least one additional predetermined grammar are each selected by the recipient of the message from a group consisting of:

a telephone number grammar, an address grammar, a date grammar, and a time of day grammar.

10. A method of recognizing information embedded in a stored voice message, wherein the method comprises steps of:

a. providing the stored voice message to a speech recognizer;

b. searching the stored voice message for all occurrences of a plurality of predetermined grammar types wherein each predetermined grammar type is one of a numeric grammar, an alpha grammar and an alphanumeric grammar, and further wherein each found occurrence corresponds to a predetermined grammar;

c. identifying information contained in the predetermined grammar; and

d. outputting the information.

11. The method according to claim 10 wherein the step of outputting comprises steps of:

a. identifying a number included in the portion of the voice message; and

b. displaying the number.

12. The method according to claim 10 wherein the voice message is stored in a telephone voice mail system.

13. The method according to claim 12 wherein the step of searching is performed while the stored voice message is being recorded by a caller.

14. The method according to claim 10 wherein the predetermined grammar is selected by a recipient of the stored voice message.

15. The method according to claim 14 wherein the recipient of message selects the predetermined grammar after listening to the stored voice message.

16. The method according to claim 14 wherein the predetermined grammar is selected from a group consisting of:

a telephone number grammar, an address grammar, a date grammar, and a time of day grammar.

17. The method according to claim 10 further comprising a step of searching for at least one additional predetermined grammar.

18. The method according to claim 17 wherein the predetermined grammar and the at least one additional predetermined grammar are each selected by the recipient of the message from a group consisting of:

a telephone number grammar, an address grammar, a date grammar, and a time of day grammar.

19. A voice recognition system for recognizing information embedded in a stored voice message, wherein the voice recognition system comprises:

a. an interface coupled to receive a voice message from a voice mail system;

b. a voice recognizer coupled to the interface to search the stored voice message for all occurrences of a plurality of predetermined grammar types wherein each predetermined grammar type is one of a numeric grammar, an alpha grammar and an alphanumeric grammar, and further wherein each found occurrence corresponds to a predetermined grammar, and to identify the predetermined grammar within the voice message; and

c. an output device coupled to the voice recognizer to provide a portion of the voice message which includes each found grammar to a user.

20. The voice recognition system according to claim 19 wherein the predetermined grammar is selected from a group consisting of:

a telephone number grammar, an address grammar, a date grammar, and a time of day grammar.

21. The voice recognition system according to claim 19 wherein the predetermined grammar includes an indicator selected from the group consisting of:

“

telephone number is”

, “

number is”

, “

phone number is”

, “

call me at”

, “

reach me at” and

“

extension”

.

22. The voice recognition system according to claim 19 wherein the voice recognizer identifies at least one additional predetermined grammar.

23. A method or recognizing information embedded in a stored voice message, wherein the method comprises steps of:

(a) searching the stored voice message for a predetermined indicator which is a grammar;

(b) searching a portion of the stored voice message adjacent to the predetermined indicator for a predetermined grammar; and

(c) recognizing the predetermined grammar.

24. The method according to claim 23 wherein the step of searching the stored voice message for the predetermined indicator comprises a step of searching the stored voice message for one or more additional predetermined indicators.

25. The method according to claim 24 wherein the predetermined grammar is selected according to which predetermined indicator is identified in the stored voice message.

26. The method according to claim 23 wherein the predetermined indicator is selected from the group of phrases consisting of:

“

telephone number is”

, “

number is”

, “

phone number is”

, “

call me at”

, “

reach me at” and

“

extension”

.

27. The method according to claim 23 wherein the stored voice message is stored in a telephone voice mail system.

28. The method according to claim 23 further comprising a step of outputting a portion of the telephone message including the predetermined grammar.

29. A method of recognizing information embedded in a stored voice message, wherein the method comprises steps of:

(a) segmenting the voice message into a plurality of segments;

(b) searching the plurality of segments for a predetermined indicator which is a grammar; and

(c) identifying a predetermined grammar adjacent to the predetermined indicator.

30. The method according to claim 29 wherein the voice message is stored in a telephone voice mail system.

31. The method according to claim 30 further comprising a step of outputting a portion of the stored voice message which includes the identified grammar.

32. The method according to claim 29 further comprising a step of outputting a portion of the stored voice message which includes the identified grammar and the identified indicator.

33. A method of recognizing information embedded in a stored voice message, wherein the method comprises steps of:

(a) segmenting the voice message into a plurality of segments;

(b) searching the plurality of segments for a predetermined indicator which is a grammar; and

(c) outputting a predetermined grammar adjacent to the predetermined indicator.

34. The method according to claim 33 wherein the stored voice message is stored in a telephone voice mail system.

35. The method according to claim 34 wherein the step of outputting comprises a step of audibly producing the predetermined length of the stored voice message.

36. The method according to claim 35 further comprising a step of outputting the predetermined indicator.

37. A method of recognizing information embedded in a stored voice message, wherein the method comprises steps of:

a. providing the stored voice message to a speech recognizer;

b. searching the stored voice message for a predetermined grammar wherein the predetermined grammar is one of a numeric grammar, an alpha grammar and an alphanumeric grammar;

c. identifying a portion of the stored voice message including the predetermined grammar; and

d. outputting the portion of the stored voice message including the predetermined grammar and not outputting substantially all of a remaining portion of the stored voice message.

38. The method according to claim 37 wherein the step of outputting comprises a step of audibly replaying the portion of the stored voice message which includes the predetermined grammar.

39. The method according to claim 37 wherein the voice message is stored in a telephone voice mail system.

40. The method according to claim 39 wherein the step of searching is performed while the stored voice message is being recorded by a caller.

41. The method according to claim 37 wherein the predetermined grammar is selected by a recipient of the stored voice message.

42. The method according to claim 41 wherein the recipient of message selects the predetermined grammar after listening to the stored voice message.

43. The method according to claim 41 wherein the predetermined grammar is selected from a group consisting of:

a telephone number grammar, an address grammar, a date grammar, a person'"'"'s name grammar and a time of day grammar.

44. The method according to claim 37 further comprising a step of searching for at least one additional predetermined grammar.

45. The method according to claim 44 wherein the predetermined grammar and the at least one additional predetermined grammar are each selected by the recipient of the message from a group consisting of:

a telephone number grammar, an address grammar, a date grammar, and a time of day grammar.