System and method for enhancing speech recognition accuracy using weighted grammars based on user profile including demographic, account, time and date information

• US 9,741,340 B2
• Filed: 11/07/2014
• Issued: 08/22/2017
• Est. Priority Date: 11/07/2014
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Czahor, Michael
• Primary Examiner(s): Kazeminezhad, Farzad
• Application Number:
  US14/535,869
• Publication Number:
  US 20160133250A1
• Time in Patent Office: 1,019 Days
• Field of Search: 704277, 704275
• CPC Class Codes: G06F 40/58 Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation
• CPC Class Codes: G10L 15/19 Grammatical context, e.g. disambiguation of the recognition hypotheses based on word sequence rules
• CPC Class Codes: G10L 2015/227 of the speaker; Human-factor methodology
• CPC Class Codes: G10L 2015/228 of application context

Abstract

Disclosed herein are systems, computer-implemented methods, and computer-readable media for enhancing speech recognition accuracy. The method includes dividing a system dialog turn into segments based on timing of probable user responses, generating a weighted grammar for each segment, exclusively activating the weighted grammar generated for a current segment of the dialog turn during the current segment of the dialog turn, and recognizing user speech received during the current segment using the activated weighted grammar generated for the current segment. The method can further include assigning probability to the weighted grammar based on historical user responses and activating each weighted grammar is based on the assigned probability. Weighted grammars can be generated based on a user profile. A weighted grammar can be generated for two or more segments. The weighted grammar is weighted based on a user profile which consists of information about a number called from, demographic information, account information, a time of day, and a date. Exclusively activating each weighted grammar can include a transition period blending the previously activated grammar and the grammar to be activated.

Claims

1. A method comprising:

identifying, via a processor, a first segment of a dialog turn associated with soliciting a first probable user response as part of a dialog with a dialog system;

identifying, via the processor, a second segment of the dialog turn associated with soliciting a second probable user response, wherein the first segment and the second segment are further identified based on a first timing of the first probable user response and a second timing of the second probable user response;

activating a first weighted grammar for the first segment of the dialog to process speech received during the first segment, to yield a first activated weighted grammar, wherein the first weighted grammar is weighted based on a user profile which consists of information about a number called from, demographic information, account information, a time of day, and a date;

activating a second weighted grammar for the second segment of the dialog to process speech received during the second segment, to yield a second activated weighted grammar;

recognizing user speech received during the first segment of the dialog using the first activated weighted grammar; and

recognizing user speech received during the second segment of the dialog using the second activated weighted grammar.

2. The method of claim 1, further comprising assigning a first probability to the first weighted grammar and a second probability to the second weighted grammar based on historical user responses and activating the first weighted grammar based on the first probability and activating the second weighted grammar based on the second probability.

3. The method of claim 1, further comprising weighting a grammar based on an expected probable user response, to yield the first weighted grammar.

4. The method of claim 1, wherein activating the first weighted grammar further comprises initiating a transition period blending a previously activated grammar and the first weighted grammar.

5. The method of claim 1, wherein two identified segments are associated with the first weighted grammar.

6. The method of claim 1, further comprising presenting a menu to a user using the dialog system prior to receiving the first user speech or the second user speech.

7. A system comprising:

a processor; and

a computer-readable storage medium having instructions stored which, when executed on the processor, cause the processor to perform operations comprising;

identifying, via a processor, a first segment of a dialog turn associated with soliciting a first probable user response as part of a dialog with a dialog system;

identifying, via the processor, a second segment of the dialog turn associated with soliciting a second probable user response, wherein the first segment and the second segment of the dialog turn are further identified based on a first timing of the first probable user response and a second timing of the second probable user response;

activating a first weighted grammar for the first segment of the dialog for processing speech received during the first segment, to yield a first activated weighted grammar, wherein the first weighted grammar is weighted based on a user profile which consists of information about a number called from, demographic information, account information, a time of day, and a date;

activating a second weighted grammar for the second segment of the dialog for processing speech received during the second segment, to yield a second activated weighted grammar;

recognizing user speech received during the first segment of the dialog using the first activated weighted grammar; and

recognizing user speech received during the second segment of the dialog using the second activated weighted grammar.

8. The system of claim 7, the computer-readable storage medium having additional instructions stored which, when executed by the processor, result in operations comprising assigning a first probability to the first weighted grammar and a second probability to the second weighted grammar based on historical user responses and activating the first weighted grammar based on the first probability and activating the second weighted grammar based on the second probability.

9. The system of claim 7, the computer-readable storage medium having additional instructions stored which, when executed by the processor, result in operations comprising weighting a grammar based on an expected probable user response to yield the first weighted grammar.

10. The system of claim 7, wherein activating the first weighted grammar further comprises initiating a transition period blending a previously activated grammar and the first weighted grammar.

11. The system of claim 7, wherein two identified segments are associated with the first weighted grammar.

12. The system of claim 7, the computer-readable storage medium having additional instructions stored which, when executed by the processor, result in operations comprising presenting a menu to a user using the dialog system prior to receiving the first user speech or the second user speech.

13. A computer-readable storage device having instructions stored which, when executed on a computing device, cause the computing device to perform operations comprising:

identifying, via a processor, a first segment of a dialog turn associated with soliciting a first probable user response as part of a dialog with a dialog system;

identifying, via the processor, a second segment of the dialog turn associated with soliciting a second probable user response, wherein the first segment and the second segment are further identified based on a first timing of the first probable user response and a second timing of the second probable user response;

activating a first weighted grammar for the first segment of the dialog for processing speech received during the first segment, to yield a first activated weighted grammar, wherein the first weighted grammar is weighted based on a user profile which consists of information about a number called from, demographic information, account information, a time of day, and a date;

activating a second weighted grammar for the second segment of the dialog for processing speech received during the second segment, to yield a second activated weighted grammar;

recognizing user speech received during the first segment of the dialog using the first activated weighted grammar; and

recognizing user speech received during the second segment of the dialog using the second activated weighted grammar.

14. The computer-readable storage device of claim 13, having additional instructions stored which, when executed by the computing device, result in operations comprising assigning a first probability to the first weighted grammar and a second probability to the second weighted grammar based on historical user responses and activating the first weighted grammar based on the first probability and activating the second weighted grammar based on the second probability.

15. The computer-readable storage device of claim 13, having additional instructions stored which, when executed by the computing device, result in operations comprising weighting a grammar based on an expected probable user response, to yield the first weighted grammar.