COMMAND BOUNDARY IDENTIFIER FOR CONVERSATIONAL NATURAL LANGUAGE

• US 20010056344A1
• Filed: 10/28/1998
• Published: 12/27/2001
• Est. Priority Date: 10/28/1998
• Status: Active Grant

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: International Business Machines Corporation
• Inventors: KLEINDIENST, JAN, RAMASWAMY, GANESH N.
• Granted Patent: US 6,453,292 B2
• Time in Patent Office: Days
• US Class Current: 704/235
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/088 Word spotting
• CPC Class Codes: G10L 2015/223 Execution procedure of a spoken command

Abstract

An apparatus for automatically identifying command boundaries in a conversational natural language system, in accordance with the present invention, includes a speech recognizer for converting an input signal to recognized text and a boundary identifier coupled to the speech recognizer for receiving the recognized text and determining if a command is present in the recognized text, the boundary identifier outputting the command if present in the recognized text. A method for identifying command boundaries in a conversational natural language system is also included.

Claims

1. An apparatus for automatically identifying command boundaries in a conversational natural language system, comprising:

a speech recognizer for converting an input signal to recognized text; and

a boundary identifier coupled to the speech recognizer for receiving the recognized text and determining if a command is present in the recognized text, the boundary identifier outputting the command if present in the recognized text.

2. The apparatus as recited in

claim 1

, wherein the boundary identifier outputs to an application which executes the command.

3. The apparatus as recited in

claim 1

, wherein the boundary identifier includes an input processor for processing the recognized text.

4. The apparatus as recited in

claim 3

, wherein the input processor processes the recognized text by augmenting each word in the recognized text by the word'"'"'s relative position with respect to a hypothesized command boundary.

5. The apparatus as recited in

claim 3

, wherein the boundary identifier further comprises a feature detector coupled to the input processor, the feature detector for determining which feature functions are present, from a set of feature functions, in the processed recognized text.

6. The apparatus as recited in

claim 5

, wherein the boundary identifier further comprises a decision maker for determining if a command is present in the processed recognized text according to a set of feature weights corresponding to the feature functions in the processed recognized text.

7. The apparatus defined in

claim 6

, wherein the decision maker is coupled to the feature detector and decides if the processed recognized text includes a command boundary.

8. The apparatus as recited in

claim 1

, further comprising a training system for training the apparatus to recognize text and to recognize complete commands.

9. The apparatus as recited in

claim 8

, wherein the training system includes an input processor for processing a collection of training data comprising utterances which include complete commands and other than complete commands.

10. The apparatus as recited in

claim 9

, wherein the input processor inserts a token before each utterance in the training data.

11. The apparatus as recited in

claim 9

, wherein the input processor inserts a token before a first utterance in the recognized text, and after every command in the recognized text.

12. The apparatus as recited in

claim 8

, further comprises a feature extractor for extracting feature functions including words and relative positions of the words with respect to a hypothesized boundary location.

13. The apparatus as recited in

claim 1

, wherein the speech recognizer includes a language model that has been trained using training data, the training data including a token inserted to indicate a location of a command boundary in the training data.

14. The apparatus as recited in

claim 13

, wherein the speech recognizer includes additional baseforms for the token.

15. The apparatus as recited in

claim 14

, wherein the speech recognizer produces the recognized text including the token.

16. The apparatus as recited in

claim 1

, wherein the boundary identifier declares a command boundary when there is an extended period of silence in the recognized text.

17. A program storage device readable by machine, tangibly embodying a program of instructions executable by the machine to perform method steps for identifying commands in recognized text, the method steps comprising:

inputting recognized text;

processing the recognized text by augmenting words of the recognized text with a position relative to a hypothesized command boundary;

determining feature functions in the processed recognized text in accordance with a set of feature functions;

deciding whether the processed recognized text with feature functions identified includes a command, the decision being made base on weighting of feature functions; and

if a command is included, outputting the command.

18. The program storage device as recited in

claim 17

, further comprises a program of instructions for training the program storage device by inputting training data including utterances comprising commands and other than commands.

19. The program storage device as recited in

claim 18

, wherein the program of instructions for training the program storage device includes the step of placing a token before each utterance.

20. The program storage device as recited in

claim 18

, wherein the program of instructions for training the program storage device includes the step of placing a token after each command boundary included in the utterances.

21. The program storage device as recited in

claim 17

, wherein the program of instructions for training the program storage device includes the step of extracting feature functions from the training data.

22. The program storage device as recited in

claim 17

, wherein the program of instructions for training the program storage device includes the step of determining feature weights for all feature functions.

23. The program storage device as recited in

claim 17

, wherein the program of instructions for processing the recognized text includes the step of placing a token before a first utterance in the recognized text and after each command in the recognized text.

24. The program storage device as recited in

claim 17

, further includes a speech recognizer for providing the recognized text.

25. A method for identifying command in natural conversational language comprising the steps of:

inputting recognized text;

processing the recognized text by augmenting words of the recognized text with a position relative to a hypothesized command boundary;

determining feature functions in the processed recognized text in accordance with a set of feature functions;

deciding whether the processed recognized text with feature functions identified includes a command, the decision being made base on weighting of feature functions; and

if a command is included, outputting the command.

26. The method as recited in

claim 25

, further comprises the step of inputting training data including utterances comprising commands and other than commands to an input processor.

27. The method as recited in

claim 25

, further comprises the step of placing a token before each utterance of the training data.

28. The method as recited in

claim 26

, further comprises the step of placing a token after command boundaries included in the utterances.

29. The method as recited in

claim 26

, further comprises the step of extracting feature functions from the training data.

30. The method as recited in

claim 26

, further comprises the step of determining feature weights for all feature functions.

31. The method as recited in

claim 26

, further comprises the step of placing a token before a first utterance in the recognized text and after each command in the recognized text.

32. The method as recited in

claim 25

, further comprises the step of outputting the command to a device for executing the command includes a speech recognizer for providing the recognized text.