Mass-scale, user-independent, device-independent voice messaging system

• US 8,934,611 B2
• Filed: 10/31/2007
• Issued: 01/13/2015
• Est. Priority Date: 02/10/2006
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Doulton, Daniel Michael
• Primary Examiner(s): King, Simon
• Application Number:
  US11/931,000
• Publication Number:
  US 20080052070A1
• Time in Patent Office: 2,631 Days
• Field of Search: 704/231, 704/215, 704/9, 704/270, 379/88.01, 379/88.14
• US Class Current: 379/88.01
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: H04M 2201/60 Medium conversion
• CPC Class Codes: H04M 3/4936 Speech interaction details speech recognition per se G10L15/00
• CPC Class Codes: H04M 3/5183 Call or contact centers with computer-telephony arrangements
• CPC Class Codes: H04M 3/53333 Message receiving aspects

Abstract

A mass-scale, user-independent, device-independent, voice messaging system that converts unstructured voice messages into text for display on a screen is disclosed. The system comprises (i) computer implemented sub-systems and also (ii) a network connection to human operators providing transcription and quality control; the system being adapted to optimize the effectiveness of the human operators by further comprising 3 core sub-systems, namely (i) a pre-processing front end that determines an appropriate conversion strategy; (ii) one or more conversion resources; and (iii) a quality control sub-system.

Claims

1. A voice messaging system for converting an audio voice message from a caller to text, the voice messaging system comprising:

a plurality of conversion resources for converting the audio voice message into the text for an intended recipient, the plurality of conversion resources comprising;

at least one automatic speech recognition (ASR) system to automatically recognize at least some of the audio voice message to convert, at least in part, the audio voice message into the text;

a computer implemented call-pair sub-system configured to use call-pair history information based on at least one previous communication between the caller and the recipient to improve conversion accuracy of the ASR system in converting the audio voice message to the text, wherein the call-pair history information includes at least one word or phrase used by the caller or the intended recipient in a previous communication to form a grammar to improve the conversion accuracy of the ASR system in converting the audio voice message to the text; and

a text output device configured to output the text.

2. The system of claim 1 in which the call-pair history acquires, over time, without explicit user training, user-dependent data that improves the conversion accuracy.

3. The system of claim 1 in which the call-pair history information comprises at least one of the following:

a telephone number of the caller, a telephone number of the intended recipient, IP addresses, and e-mail addresses.

4. The system of claim 1 in which the call-pair history information comprises a language being used by the caller or the intended recipient.

5. The system of claim 1 in which the call-pair history information comprises a country associated with the caller or the intended recipient.

6. The system of claim 1 in which the call-pair history information comprises at least one of:

time zones, a time the audio voice message was received, and a day the audio voice message was received.

7. The system of claim 1 in which the at least one word or phrase stored in the call-pair history information includes at least one word or phrase used by the caller or the intended recipient during the previous communication between the caller and the intended recipient.

8. The system of claim 1 in which the call-pair history includes information comprises intonation information associated with the caller or the intended recipient.

9. The system of claim 1 in which the call-pair history includes information comprises personal information management data.

10. The system of claim 1 further comprising a computer implemented dynamic language model sub-system configured to build a dynamic language model using one or more of:

caller dependence;

call-pair dependence;

intended recipient dependence.

11. The system of claim 1 in which the audio voice message is a voicemail intended for a mobile telephone and the audio voice message is converted to text and sent to that mobile telephone via the text output device.

12. The system of claim 1 in which the audio voice message is intended for an instant messaging service and the audio voice message is converted to text and sent to an instant messaging service for a display on a screen via the text output device.

13. The system of claim 1 in which the audio voice message is intended for a web service and the audio voice message is converted to text and sent to a server for display as part of the web service via the text output device.

14. The system of claim 1 in which the audio voice message is intended to be converted to text format and sent as a text message via the text output device.

15. The system of claim 1 in which the audio voice message is intended to be converted to text format and sent as an email message via the text output device.

16. The system of claim 1 in which the audio voice message is intended to be converted to text format and sent as a note or memo, by email or text, to an originator of the message via the text output device.

17. The system of claim 1 including a mobile telephone network over which the audio voice message is received.

18. The system of claim 1 wherein the output device is configured to out output the text to a mobile telephone over a mobile telephone network.

19. A method for converting an audio voice message from a caller, to a recipient, to text, the method comprising:

automatically recognizing, using at least one processor configured to perform automatic speech recognition, at least some of the audio voice message to convert, at least in part, the audio voice message into the text;

using, by the at least one processor, call-pair history information based on at least one previous communication between the caller and the recipient to improve conversion accuracy of the ASR system in converting the audio voice message to text, wherein the call-pair history information includes at least one word or phrase used by the caller or the recipient in a previous communication to form a grammar to improve the conversion accuracy of the ASR system in converting the audio voice message to the text; and

outputting the text.

20. The method of claim 19 further comprising building a dynamic language model using one or more of:

caller dependence;

call-pair dependence;

intended recipient dependence.

21. At least one computer readable storage device for storing instructions that, when executed on at least one computer, performs a method for converting an audio voice message from a caller, to a recipient, to text, the method comprising:

automatically recognizing at least some of the audio voice message to convert, at least in part, the audio voice message into the text;

using call-pair history information based on at least one previous communication between the caller and the recipient to improve conversion accuracy of the ASR system in converting the audio voice message to text, wherein the call-pair history information includes at least one word or phrase used by the caller or the recipient in a previous communication to form a grammar to improve the conversion accuracy of the ASR system in converting the audio voice message to the text; and

outputting the text.

22. The at least one computer readable storage device of claim 21 further comprising building a dynamic language model using one or more of:

caller dependence;

call-pair dependence;

intended recipient dependence.

23. The at least one computer readable storage device of claim 21, wherein the at least one word or phrase stored in the call-pair history information includes at least one word or phrase used by the caller or the intended recipient during the previous communication between the caller and the intended recipient.