Assigning meanings to utterances in a speech recognition system

• US 6,704,710 B2
• Filed: 10/12/2001
• Issued: 03/09/2004
• Est. Priority Date: 12/31/1992
• Status: Expired due to Fees

• Current Assignee: Apple Inc.
• Original Assignee: Apple Computer Incorporated (Apple Inc.)
• Inventors: Strong, Robert Don
• Primary Examiner(s): MCFADDEN, SUSAN IRIS
• Application Number:
  US09/977,520
• Publication Number:
  US 20020016710A1
• Time in Patent Office: 879 Days
• Field of Search: 704/200, 704/231, 704/243, 704/244, 704/251, 704/255, 704/257
• US Class Current: 704/255
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding
• CPC Class Codes: G10L 2015/228 of application context

Abstract

Assigning meanings to spoken utterances in a speech recognition system. A plurality of speech rules is generated, each of the of speech rules comprising a language model and an expression associated with the language model. At one interval (e.g. upon the detection of speech in the system), a current language model is generated from each language model in the speech rules for use by a recognizer. When a sequence of words is received from the recognizer, a set of speech rules which match the sequence of words received from the recognizer is determined. Each expression associated with the language model in each of the set of speech rules is evaluated, and actions are performed in the system according to the expressions associated with each language model in the set of speech rules.

Claims

1. A machine-readable medium having stored thereon a sequence of instructions which when executed by a processing system cause said processing system to perform a method of speech recognition in a speech recognition system, the method comprising:

storing a language model defining the syntax of a set of sequences of words which may be recognized in each of a plurality of speech rules;

storing an expression defining the meaning of each of said set of sequences of words in each of said plurality of speech rules;

providing said language model to a recognizer and responsive thereto, receiving a recognized sequence of words from said recognizer;

determining a set of said plurality of speech rules which include language models matching said recognized sequence of words; and

evaluating said expression for each of said set of said plurality of speech rules and performing actions according to each said expression which is evaluated.

2. The machine-readable medium of claim 1 wherein each said expression comprises an arithmetic calculation for variables in said speech recognition system.

3. The machine-readable medium of claim 1 wherein said expression specifies a category.

4. The machine-readable medium of claim 1 wherein said expression specifies a command.

5. The machine-readable medium of claim 1 wherein certain of said speech rules comprise recursive references to others of said plurality of speech rules.

6. The machine-readable medium of claim 5 wherein said evaluation comprises evaluating each of said expressions for said other language models prior to evaluating each said language model in each speech rule of said set of speech rules.

7. A machine-readable medium having stored thereon a sequence of instructions which when executed by a processing system cause said processing system to perform a method of associating meanings to utterances in a speech recognition system, the method comprising:

providing a plurality of speech rules, each speech rule comprising a language model and an expression defining a meaning of said speech rule;

generating a current language model from each said language model of said plurality of speech rules and providing said current language model to a recognizer;

said recognizer recognizing words in detected speech by referencing said current language model to generate a recognized sequence of words;

receiving said recognized sequence of words from said recognizer, and determining that said recognized sequence of words matches a phrase of a first speech rule of said plurality of speech rules and that said recognized sequence of words comprises at least one word that matches a phrase of a second speech rule of said plurality of speech rules; and

evaluating a first expression of the first speech rule and a second expression of the second speech rule, wherein the evaluation of the first expression depends on the evaluation of the second expression, and performing actions in said speech recognition system only after evaluating the first and second expressions.

8. The machine-readable medium of claim 7 wherein each said expression comprises an arithmetic calculation for said variables in said speech recognition system.

9. The machine-readable medium of claim 7 wherein each said language model in each of said plurality of speech rules comprises recursive references to other language models.

10. The machine-readable medium of claim 9 wherein each of said other language models comprise expressions associated with each of said other language models.

11. The machine-readable medium of claim 10 wherein said evaluating a first expression comprises evaluating each of said expressions for said other language models prior to evaluating each said language model in each of said first and second speech rules.

12. A machine-readable medium having stored thereon a sequence of instructions which when executed by a processing system cause said processing system to perform a method of speech recognition in a speech recognition system, the method comprising:

storing a plurality of speech rules, each of said plurality of speech rules including a language model defining a set of sequences of words which may be recognized, and an expression defining a meaning of said set of sequences of words;

generating a current language model including a first set of language models from said plurality of said speech rules, and providing said current language model to a recognizer for recognition of utterances;

said recognizer recognizing words contained in an utterance accordance to said current language model, and generating a recognized sequence of words;

determining a matched set of speech rules according to said recognized sequence of words; and

evaluating each said expression in each of said set of matched speech rules, and performing actions in said system according to each said expression.

13. The machine-readable medium of claim 12 wherein each said expression comprises an arithmetic calculation for variables in said speech recognition system.

14. The machine-readable medium of claim 12 wherein said expression specifies a category.

15. The method of claim 12 wherein said expression specifies a command.

16. The method of claim 12 wherein certain of said speech rules comprise recursive references to others of said plurality of speech rules.

17. The method of claim 15 wherein evaluating each said expression comprises evaluating each of said expressions for said other language models prior to evaluating each said language model in each speech rule of said set of speech rules.

18. An apparatus for speech recognition in a speech recognition system comprising:

means for storing a plurality of speech rules, each of said plurality of speech rules including a language model defining a set of sequences of words which may be recognized, and an expression defining a meaning of said set of sequences of words;

means for generating a current language model including a first set of language models from said plurality of said speech rules, and providing said current language model to a recognizer for recognition of utterances;

means for said recognizer to recognize words contained in an utterance according to said current language model to generate a recognized sequence of words;

means for determining a matched set of speech rules according to said recognized sequence of words; and

means for evaluating each said expression in each of said set matched speech rules, and to perform actions in said system according to each said expression.

19. An apparatus for speech recognition in a speech recognition system comprising:

means for storing a language model defining the syntax of a set of sequences of words which may be recognized in each of a plurality of speech rules;

means for storing an expression defining the meaning of each of said set of sequences of words in each of said plurality of speech rules;

means for providing said language model to a recognizer and in response thereto to receive a recognized sequence of words from said recognizer;

means for determining a set of said plurality of speech rules which include language models matching said recognized sequence of rules; and

means for evaluating said expression for each of said set of said plurality of speech rules and performing actions according to each said expression which is evaluated.