Sentence simplification for spoken language understanding

• US 9,454,962 B2
• Filed: 05/12/2011
• Issued: 09/27/2016
• Est. Priority Date: 05/12/2011
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Tur, Gokhan, Hakkani-Tur, Dilek, Heck, Larry Paul, Parthasarathy, Sarangarajan
• Primary Examiner(s): Sirjani, Fariba
• Application Number:
  US13/106,374
• Publication Number:
  US 20120290290A1
• Time in Patent Office: 1,965 Days
• Field of Search: 704/9
• US Class Current: 1/1
• CPC Class Codes: G06F 40/211 Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
• CPC Class Codes: G06F 40/253 Grammatical analysis; Style critique
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2015/223 Execution procedure of a spoken command

Abstract

Sentence simplification may be provided. A spoken phrase may be received and converted to a text phrase. An intent associated with the text phrase may be identified. The text phrase may then be reformatted according to the identified intent and a task may be performed according to the reformatted text phrase.

Claims

1. A computer-implemented method being performed at a computing device, the method comprising:

receiving a spoken utterance at the computing device;

converting the spoken utterance to a text phrase;

identifying a top level predicate of the text phrase, wherein the top level predicate comprises a word of the text phrase upon which other words of the text phrase depend and which does not depend on other words of the text phrase, wherein the top level predicate is associated with an intent of the text phrase;

reformatting the text phrase, wherein the reformatted text phrase comprises at least the word corresponding to the top level predicate;

defining an intent domain based on the reformatted text phrase;

filling one or more semantic slots of a spoken language understanding application associated with the intent domain with words from the text phrase; and

performing a task associated with the intent domain and based on information from the filled semantic slots of the spoken language understanding application at the computing device.

2. The method of claim 1, wherein identifying the top level predicate of the text phrase comprises performing a dependency parse on the text phrase.

3. The method of claim 2, wherein performing a dependency parse comprises:

identifying a top level predicate; and

excluding at least one auxiliary word in the text phrase.

4. The method of claim 3, wherein the at least one auxiliary word comprises a dependent of the top level predicate.

5. The method of claim 3, wherein the at least one auxiliary word comprises at least one predefined auxiliary keyword.

6. The method of claim 3, wherein identifying the top level predicate comprises evaluating a weighting criterion associated with each word of the text phrase.

7. The method of claim 1, wherein reformatting the text phrase further comprises excluding a word of the text phrase.

8. The method of claim 1, wherein filling one or more semantic slots further comprises filling a slot with an implied word.

9. The method of claim 1, wherein filling one or more semantic slots is performed based in part on the top level predicate and one or more predicates of other clauses of the text phrase.

10. The method of claim 9, wherein the one or more predicates of other clauses are determined recursively.

11. A computer storage device which stores a set of instructions which when executed by a processor performs a method for providing sentence simplification for spoken language understanding, the method executed by the set of instructions comprising:

identifying a predicate associated with a text phrase, wherein the predicate comprises a word of the text phrase upon which other words of the text phrase depend and which does not depend on other words of the text phrase, wherein the predicate is associated with an intent of the text phrase;

reformatting the text phrase, wherein the reformatted text phrase comprises at least the word corresponding to the predicate;

defining an intent domain based on the reformatted text phrase;

filling one or more semantic slots of a spoken language understanding application associated with the intent domain with words from the text phrase;

performing a task associated with the defined intent domain and based on information from the filled semantic slots of the spoken language understanding application at the computing device; and

displaying a result associated with the task.

12. The computer storage device of claim 11, wherein the text phrase comprises a conversion of a spoken phrase received from a user.

13. The computer storage device of claim 11, wherein the text phrase is associated with a web search query and identifying the predicate comprises defining a target domain.

14. The computer storage device of claim 13, wherein the reformatted text phrase comprises at least one keyword associated with the target domain.

15. The computer storage device of claim 14, wherein the target domain comprises a plurality of semantic slots.

16. The computer storage device of claim 11, wherein filling one or more semantic slots further comprises filling a slot with an implied word.

17. The computer storage device of claim 11, wherein filling one or more semantic slots is performed based in part on the top level predicate and one or more predicates of other clauses of the text phrase.

18. The computer storage device of claim 11, wherein identifying the predicate comprises performing a dependency parse on the text phrase.

19. The computer storage device of claim 18, wherein identifying the predicate is determined recursively.

20. A system for providing sentence simplification for spoken language understanding, the system comprising:

a memory storage; and

a processing unit coupled to the memory storage, wherein the processing unit is operative to;

receive a spoken phrase;

convert the spoken phrase to a text phrase;

perform a dependency parse on the text phrase;

identify a word corresponding to a top level predicate of the text phrase, wherein the top level predicate comprises a word of the text phrase upon which other words of the text phrase depend and which does not depend on other words of the text phrase, wherein the top level predicate is associated with an intent of the text phrase;

reformat the text phrase, wherein the reformatted text phrase comprises at least the word corresponding to the top level predicate;

define an intent domain based on the reformatted text phrase, wherein the defined intent domain is associated with a spoken language understanding application which comprises a plurality of semantic slots;

fill the plurality of semantic slots of the spoken language understanding application with words from the text phrase;

perform a task associated with the intent domain and based on the information from the filled plurality of semantic slots of the spoken language understanding application; and

display a result of the performed task to a user.