System and method for improving speech recognition accuracy using textual context

Patent metadata

Field	Value
Publication number	US-10546595-B2
Application number	US-201815911678-A
Country	US
Kind code	B2
Filing date	Mar 5, 2018
Priority date	Oct 23, 2009
Publication date	Jan 28, 2020
Grant date	Jan 28, 2020

Abstract

Disclosed herein are systems, methods, and computer-readable storage media for improving speech recognition accuracy using textual context. The method includes retrieving a recorded utterance, capturing text from a device display associated with the spoken dialog and viewed by one party to the recorded utterance, and identifying words in the captured text that are relevant to the recorded utterance. The method further includes adding the identified words to a dynamic language model, and recognizing the recorded utterance using the dynamic language model. The recorded utterance can be a spoken dialog. A time stamp can be assigned to each identified word. The method can include adding identified words to and/or removing identified words from the dynamic language model based on their respective time stamps. A screen scraper can capture text from the device display associated with the recorded utterance. The device display can contain customer service data.

First claim

What is claimed is:

1. A method comprising:
identifying words with a speech recognition model to yield identified words;
removing a word of the identified words from the speech recognition model according to at least an end time stamp assigned to the word to yield a modified speech recognition model; and
recognizing an utterance using the modified speech recognition model.

2. The method of claim 1 , wherein the speech recognition model comprises a first recognition model that is modified by associating the identified words to yield the speech recognition model.

3. The method of claim 2 , wherein each word in the identified words is assigned a beginning time stamp and an end time stamp, and wherein the associating of the identified words comprises adding each word to the speech recognition model according to the beginning time stamp assigned to each word.

4. The method of claim 1 , wherein the identified words comprise information associated with a product or a service.

5. The method of claim 4 , wherein the identified words are presented on a display of an agent interacting with a user to assist in a purchase of the product or service.

6. The method of claim 1 , wherein the identified words are associated with social media content.

7. The method of claim 1 , wherein the identified words represent a portion of a spoken dialog.

8. A system comprising:
a processing system including a processor; and
a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations, comprising:
identifying words with a speech recognition model to yield identified words;
removing a word of the identified words from the speech recognition model according to at least an end time stamp assigned to the word to yield a modified speech recognition model; and
recognizing an utterance using the modified speech recognition model.

9. The system of claim 8 , wherein the speech recognition model comprises a first recognition model that is modified by associating the identified words to yield the speech recognition model.

10. The system of claim 9 , wherein each word in the identified words is assigned a beginning time stamp and an end time stamp, and wherein the associating of the identified words comprises adding each word to the speech recognition model according to the beginning time stamp assigned to each word.

11. The system of claim 8 , wherein the identified words comprise information associated with a product or a service.

12. The system of claim 11 , wherein the identified words are presented on a display of an agent interacting with a user to assist in a purchase of the product or service.

13. The system of claim 8 , wherein the identified words are associated with social media content.

14. The system of claim 8 , wherein the identified words represent a portion of a spoken dialog.

15. A non-transitory machine-readable storage device, wherein the non-transitory machine-readable storage device comprises executable instructions which, when executed by a processor, cause the processor to perform operations comprising:
identifying words with a speech recognition model to yield identified words;
removing a word of the identified words from the speech recognition model according to at least an end time stamp assigned to the word to yield a modified speech recognition model; and
recognizing an utterance using the modified speech recognition model.

16. The non-transitory machine-readable storage device of claim 15 , wherein the speech recognition model comprises a first recognition model that is modified by associating the identified words to yield the speech recognition model.

17. The non-transitory machine-readable storage device of claim 16 , wherein each word in the identified words is assigned a beginning time stamp and an end time stamp, and wherein the associating of the identified words comprises adding each word to the speech recognition model according to the beginning time stamp assigned to each word.

18. The non-transitory machine-readable storage device of claim 16 , wherein each word in the identified words is assigned a beginning time stamp, and wherein the associating of the identified words comprises adding each word to the modified speech recognition model according to the beginning time stamp.

19. The non-transitory machine-readable storage device of claim 15 , wherein the identified words comprise information associated with a product or a service.

20. The non-transitory machine-readable storage device of claim 19 , wherein the identified words are presented on a display of an agent interacting with a user to assist in a purchase of the product or service.

21. The non-transitory machine-readable storage device of claim 15 , wherein the identified words are associated with social media content.

Assignees

• Nuance Communications Inc

Inventors

• Melamed Dan
• Bangalore Srinivas
• Johnston Michael

Classifications

• G10L25/51 Primary

  for comparison or discrimination · CPC title

• G10L17/04

  Training, enrolment or model building · CPC title

• G10L2015/228

  of application context · CPC title

• G10L15/30

  Distributed recognition, e.g. in client-server systems, for mobile phones or network applications · CPC title

• G06F3/162

  Interface to dedicated audio devices, e.g. audio drivers, interface to CODECs · CPC title