System and method for independently recognizing and selecting actions and objects in a speech recognition system

• US 20060161431A1
• Filed: 01/14/2005
• Published: 07/20/2006
• Est. Priority Date: 01/14/2005
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Inventors: Knott, Benjamin Anthony, Bushey, Robert R., Martin, John Mills, Potvin, Carl, Sabourin, Michael
• Granted Patent: US 7,627,096 B2
• Time in Patent Office: Days
• US Class Current: 704/235
• CPC Class Codes: G10L 15/08 Speech classification or search
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding
• CPC Class Codes: G10L 2015/0631 Creating reference templates; Clustering

Abstract

A method for processing a call is disclosed. The method receives a speech input via a call and transforms at the speech input into a textual format. The method also creates a list of salient terms of actions and objects from the text, adjusts the confidence level of objects on the list if the dominant term is an action and selects a complimentary object from the list to combine with the action to form an action-object pair. The method further adjusts a confidence level of actions on the list if the dominant term is an object and selects a complementary action from the list to combine with the action to form the action-object pair, and routes the call based on the action-object pair

Claims

1. A method for processing a call comprising:

receiving a speech input via a call;

transforming the speech input into text;

creating a list of salient terms comprising actions and objects from the text, wherein the list has a dominant term;

adjusting a confidence level of objects on the list if the dominant term is an action and selecting a complimentary object from the list to combine with the action to form an action-object pair

adjusting a confidence level of actions on the list if the dominant term is an object and

selecting a complementary action from the list to combine with the object to form the action-object pair; and

routing the call based on the action-object pair.

2. The method of claim 1, further comprising masking actions and invalid objects if the dominant term is an action.

3. The method of claim 1, further comprising masking objects and invalid actions if the dominant term is an object.

4. The method of claim 1, further comprising parsing the text into an action list and an object list.

5. The method of claim 1, further comprising assigning an initial confidence level to each of the actions and the objects contained in the list of salient terms.

6. The method of claim 1, further comprising adjusting the confidence level of objects to a value that represents a probability that a given object represents an intent of the caller.

7. The method of claim 1, further comprising adjusting the confidence level of actions to a value that represents a probability that a given action represents an intent of the caller.

8. The method of claim 1, wherein the action is selected from a group consisting of a verb and an adverb-verb combination.

9. The method of claim 1, wherein the object is selected from a group consisting of a noun and an adjective-noun combination.

10. A system for routing calls comprising:

a call routing system having an interface, the call routing system configured

to receive a call;

to transform a speech signal received via the call into text; and

+P2

to create a list of salient terms containing actions and objects from the text, wherein the list contains dominant terms;

a synonym module associated with the call routing system and operable to adjust a confidence level of invalid objects in the list when there is a dominant action and to adjust a confidence level of invalid actions when there is a dominant object;

a pairing module associated with the call routing system and operable to select a complementary action from the list when there is a dominant object and a complementary object when there is a dominant action, to create an action-object pair; and

a switch to route the call based on the action-object pair.

11. The system of claim 10, wherein the call routing system uses phonemes to convert the speech input to a word string.

12. The system of claim 11, wherein the call routing system assigns a confidence value to the word string.

13. The system of claim 11, wherein the call routing system parses the word string into a respective action and a respective object.

14. The system of claim 11, wherein the call routing system assigns a confidence value to the word string and the confidence value represents a probability that the word string represents an intent of the caller.

15. The system of claim 10, wherein the call routing system assigns a confidence value to the object and the confidence value represents a probability that the object represents an intent of the caller.

16. The system of claim 10, wherein the call routing system assigns a confidence value to the action and the confidence value represents a probability that the action represents an intent of the caller.

17. The system of claim 10, wherein the action is one of a verb or an adverb-verb combination.

18. The system of claim 10, wherein the object is one of a noun or an adjective-noun combination.

19. The system of claim 10, wherein the call routing system is operable to route the call to a destination.

20. A system comprising:

an acoustic engine configured to accept a speech input and to produce a textual version of at least a portion of the speech input as its output;

a semantic engine coupled to the acoustic engine and operable to identify an action and an object indicated by the textual version;

a probability system operable to assign confidence levels to the action and the object; and

an action-object routing table operable to provide a routing destination based at least partially on the confidence levels assigned to the action and the object.

21. The system of claim 20 further comprising a plurality of different speech inputs.

22. The system of claim 20 wherein the action is selected from a list of expected utterances comprising acquire, cancel, change, inquire, inform, and how to use.

23. The system of claim 20 wherein the object is selected from a list of expected utterances comprising DSL, basic service, call notes, caller ID, bill payment, other providers, coupon specials, names and number, and store locations

24. The system of claim 20 further comprising memory storing a library that comprises a plurality of expected actions.

25. The system of claim 20 further comprising a tuning module operable to accept an input intended to improve a recognition rate of the acoustic engine.