Method for automatic speech recognition in telephony

• US 5,822,727 A
• Filed: 12/21/1995
• Issued: 10/13/1998
• Est. Priority Date: 03/30/1995
• Status: Expired due to Term

• Current Assignee: AT&T Corporation (AT&T, Inc.)
• Original Assignee: AT&T Corporation (AT&T, Inc.)
• Inventors: Garberg, Roger Borgan, Yudkowsky, M.
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): EDOUARD, PATRICK NESTOR
• Application Number:
  US08/576,740
• Time in Patent Office: 1,027 Days
• Field of Search: 395/2.52, 395/2.55, 395/2.6, 395/2.61, 395/2.79, 395/2.84, 379/59, 379/67, 379/88, 379/89
• US Class Current: 704/270.1
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: H04M 2201/40 using speech recognition
• CPC Class Codes: H04M 2201/60 Medium conversion
• CPC Class Codes: H04M 2242/22 Automatic class or number identification arrangements
• CPC Class Codes: H04M 3/382 using authorisation codes or passwords
• CPC Class Codes: H04M 3/385 using speech signals
• CPC Class Codes: H04M 3/42059 Making use of the calling party identifier
• CPC Class Codes: H04M 3/42204 Arrangements at the exchange for service or number selection by voice at the terminal H04M1/27
• CPC Class Codes: H04M 3/493 Interactive information services, e.g. directory enquiries ; Arrangements therefor, e.g. interactive voice response [IVR] systems or voice portals
• CPC Class Codes: H04M 3/4931 Directory assistance systems
• CPC Class Codes: H04Q 3/42 Circuit arrangements for indirect selecting controlled by common circuits, e.g. register controller, marker
• CPC Class Codes: H04Q 3/72 Finding out and indicating number of calling subscriber

Abstract

Disclosed is a method and system for utilizing supplemental data to enable a voice-driven telecommunication services system to build a speech recognition template data base. Such supplemental data includes textual name data associated with many or all of a telecommunication subscriber'"'"'s incoming and outgoing calls. The textual name data, which typically represents the calling or called party'"'"'s name, is used to generate a speech recognition template using text-to-speech methods. The number and the speech recognition template may then be recorded in a data base to facilitate voice-driven services in connection with that party. A method according to the present invention includes obtaining a first input, where the first input identifies an aspect of a telecommunication network user. The first input may, for example, be the telephone number of the telecommunication network user. Then, supplemental information that is related to the first input in a supplemental data base is identified and retrieved from the supplemental data base. The supplemental information may, for example, be the textual name data associated with the telecommunication network user. This supplemental information is then utilized to create a speech recognition template. The speech recognition template may thereafter be used to recognize a second input, said second input comprising a spoken utterance and containing information representative of the supplemental information.

Claims

1. In a system including an automatic speech recognition (ASR) system, a method of utilizing information stored in a supplemental data base to enhance a capability of the ASR system, the method comprising:

(a) obtaining a first input, said first input comprising a connection address associated with a telecommunication network user;

(b) identifying and retrieving supplemental information stored in the supplemental data base, the supplemental information comprising textual information related to the first input;

(c) creating a speech recognition templated derived from the supplemental information retrieved from the supplemental data base; and

(d) using the speech recognition templated to recognize a second input, said second input comprising a spoken utterance, said second input containing information representative of the supplemental information.

2. The method of claim 1 wherein step a) further comprises obtaining incoming caller line identification information from public switched telephone network equipment corresponding to an incoming call from the telecommunications network user.

3. The method of claim 1 wherein step a) further comprises obtaining a dialed telephone number in an outgoing call to the telecommunications network user.

4. The method of claim 1 wherein step a) further comprises obtaining a spoken utterance and recognizing the spoken utterance to produce the first input.

5. The method of claim 1 wherein the supplemental information comprises textual information and step c) further comprises the step of using a text-to-speech system to generate a subword transcription of the supplemental information.

6. The method of claim 1 further comprising a step of creating a record for a subscriber data base, said record including the first input and the speech recognition template.

7. The method of claim 1 further comprising a step e) performing a communication service involving the telecommunication network user after recognizing the second input.

8. The method of claim 7 wherein step e) further comprises retrieving a recorded message generated by the telecommunications network user.

9. The method of claim 7 wherein step e) further comprises automatically dialing a telephone number associated with the telecommunications network user.

10. In a telephone network, a method of using a supplemental data base in connection with an automatic speech recognition (ASR) system to enhance the capability of the telephone network, the method comprising the steps of:

(a) receiving an incoming call from a caller;

(b) obtaining the telephone number associated with the caller;

(c) retrieving supplemental information stored in a supplemental data base using the telephone number as an index, wherein the supplemental information is textual data;

(d) creating a speech recognition template corresponding to the textual data as retrieved from the supplemental data base;

(e) storing the speech recognition template and the telephone number associated with the caller; and

(f) recognizing a second utterance using a speech recognition algorithm and the speech recognition template wherein said utterance includes information identifying the caller.

11. The method of claim 10 wherein step d) further comprises using a text-to-speech system to generate a subword transcription of the supplemental information, and said subword transcription comprises the speech recognition template.

12. A voice-driven services system for use in connection with a telecommunication network, the system comprising:

a) an automatic speech recognition circuit operable to recognize an utterance by comparing the utterance to one or more speech recognition templates, said automatic speech recognition circuit including a text-to-speech system; and

b) a computer processing unit coupled to the automatic speech recognition circuit and operable to obtain a first input, said first input identifying an aspect of a telecommunication network user, and the computer processing unit is further operable to retrieve textual information from a first data base related to the first input,

and wherein the text-to-speech system is operable to generate a speech recognition template from the textual information.

13. The voice-driven services system of claim 12 wherein the automatic speech recognition circuit is further operable to recognize a second input, said second input comprising a spoken utterance containing information representative of the textual information.

14. The voice-driven services system of claim 12 wherein the computer processing unit is further operable to retrieve the first input wherein the first input comprises a connection address associated with the telecommunications network user.

15. The voice-driven services system of claim 14 wherein the computer processing unit is further operable to retrieve the first input wherein the first input comprises a connection address including a telephone number associated with the telecommunications network user.

16. The voice-driven services system of claim 15 wherein the computer processing unit is further operable to retrieve textual information from the first data base wherein the first data base contains information associating textual name information with telephone number information for a plurality of telecommunications network users.

17. The voice-driven services system of claim 12 wherein the computer processing unit is further operable to generate a subscriber data base record comprising the textual information and the speech recognition template.

18. The voice-driven services system of claim 17 wherein the computer processing unit is further operable to perform a plurality of voice-driven services including voice-dialing using the subscriber data base record.

19. The voice-driven services system of claim 12 wherein said automatic speech recognition circuit further comprises a speech generator, said speech generator operable to generate a speech output signal representative of the textual information.

20. The voice-driven services system of claim 19 wherein the computer processing unit is further operable to cause the speech generator to provide the speech output signal representative of the textual information to a subscriber line when the subscriber line is maintaining a telephone call.

21. A voice-driven services system for use in connection with a telecommunication network, the system comprising

a) a memory containing a program for execution by a computer processing unit;

and

b) a central processing unit ("CPU") operably connected to the memory to execute the program therein, wherein the CPU executes the program in the memory in order to

obtain a first input, said first input identifying an aspect of a telecommunication network user,

retrieve supplemental data related to the first input, said supplemental data comprising textual information,

direct an automatic speech recognition circuit to generate a speech recognition template from the supplemental data, and

direct the automatic speech recognition circuit to recognize a second input comprising an utterance using the speech recognition template.

22. The voice-driven services system of claim 21 further comprising an automatic speech recognition system operable to generate the speech recognition template from the supplemental data and recognize the second input using the speech recognition template.

23. The voice-driven services system of claim 21 wherein the CPU is further operable to execute the program in the memory to obtain the first input in the form of a connection address of the telecommunication network user.

24. The voice-driven services system of claim 23 wherein the CPU is further operable to execute the program in the memory to perform voice-dialing services using the recognized second input.

25. The voice-driven services system of claim 23 wherein the CPU is further operable to execute the program in the memory to generate a data base record that includes the connection address of the telecommunication network user and the generated speech recognition template.

26. The voice-driven services system of claim 25 wherein the CPU is further operable to generate and maintain a data base containing a plurality of records, each record corresponding to a telecommunications network user and including a connection address associated with the telecommunication network user and a speech recognition template associated with the telecommunication network user.

27. The voice-driven services system of claim 21 further comprising a play/record circuit operably connected to telecommunication network and the CPU, said play/record circuit operable to play recorded announcements to the telecommunication network user and record utterances spoken by the telecommunications network user.

28. The voice-driven services system of claim 27 wherein the CPU is further operable to:

direct the play/record circuit to play a recorded announcement prompting the telecommunication network user to leave a message;

direct the play/record circuit to record an utterance from the telecommunications network user; and

associate the recorded utterance to the speech recognition template.

29. The voice-driven services system of claim 28 wherein the CPU is further operable to direct the ASR circuit to recognize a second utterance provided by a subscriber using the speech recognition template and,

direct the play/record circuit to play the recorded utterance from the telecommunications network user if the recognized second utterance corresponds to the speech recognition template.

30. The voice-driven services system of claim 21 wherein the CPU is further operable to execute the program in the memory in order to obtain the first input during an incoming call to a subscriber from the telecommunication network user.

31. The voice-driven services system of claim 30 wherein the CPU is further operable to execute the program in the memory to obtain the first input wherein the first input is a telephone number of the telecommunication network user.

32. The voice-driven services system of claim 31 wherein the CPU is further operable to execute the program in the memory to obtain the telephone number of the telecommunication network user using data provided from an external switch.