Automatic grammar generation using distributedly collected knowledge

• US 7,957,968 B2
• Filed: 12/12/2006
• Issued: 06/07/2011
• Est. Priority Date: 12/22/2005
• Status: Active Grant

• Current Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Original Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Inventors: Gupta, Rakesh, Hennacy, Ken
• Primary Examiner(s): Dorvil; Richemond
• Assistant Examiner(s): GODBOLD, DOUGLAS
• Application Number:
  US11/609,683
• Publication Number:
  US 20070179777A1
• Time in Patent Office: 1,638 Days
• Field of Search: 704243-245, 707765-768
• US Class Current: 704/243
• CPC Class Codes: G06F 40/237 Lexical tools

Abstract

The invention includes a computer based system or method for automatically generating a grammar associated with a first task comprising the steps of: receiving first data representing the first task based from responses received from a distributed network; automatically tagging the first data into parts of speech to form first tagged data; identifying filler words and core words from said first tagged data; modeling sentence structure based upon said first tagged data using a first set of rules; identifying synonyms of said core words; and creating the grammar for the first task using said modeled sentence structure, first tagged data and said synonyms.

Claims

1. A computer based method for automatically generating a hierarchical grammar associated with a plurality of tasks comprising the steps of:

creating a sub-grammar for each of the plurality of tasks, wherein creating a sub-grammar for a task comprises;

receiving data representing the task based from responses received from a distributed database;

automatically tagging the data into parts of speech to form tagged data using a processor executing instructions included in a memory coupled to the processor;

identifying filler words and core words from said tagged data by applying rules to differentiate between filler words and core words using the processor executing instructions included in the memory coupled to the processor;

automatically modeling sentence structure based upon said tagged data using a set of modeling rules retrieved from the memory coupled to the processor;

automatically identifying synonyms of said core words; and

automatically creating a sub-grammar using the modeled sentence structure, the tagged data, the identified synonyms, and the identified core words; and

creating a high-level grammar by combining filler words identified in the creation of the sub-grammars.

2. The method of claim 1, further comprising the step of checking the spelling of words in said data.

3. The method of claim 1, wherein said hierarchical grammar is combined with a statistical language model to recognize speech.

4. The method of claim 1, wherein said hierarchical grammar is used to recognize speech.

5. The method of claim 1, wherein said data represents a request to perform said first task.

6. The method of claim 1, wherein said data is received using the Internet.

7. The method of claim 1, wherein said parts of speech include verbs and nouns.

8. The method of claim 1, wherein said data is received using an internet-based distributed knowledge capture system.

9. The method of claim 8, wherein said internet-based distributed knowledge capture system is an OpenMind indoor common sense knowledge capture system.

10. The method of claim 1, further comprising the step of identifying synonyms of the identified filler words.

11. The method of claim 1, wherein the high-level grammar is created concurrently with the creation of sub-grammars.

12. A computer based system for automatically generating a hierarchical grammar associated with a plurality of tasks comprising:

creation means for creating a sub-grammar for each of the plurality of tasks, wherein the creation means comprises;

receiving means for receiving data representing the task based from responses received from a distributed database;

tagging means for automatically tagging the data into parts of speech to form tagged data;

identification means for identifying filler words and core words from said tagged data by applying rules to differentiate between filler words and core words;

modeling means for automatically modeling sentence structure based upon said tagged data using a set of modeling rules;

synonym means for automatically identifying synonyms of said core words; and

grammar means for automatically creating a sub-grammar using the modeled sentence structure, the tagged data, the identified synonyms, and the identified core words; and

creation means for creating a high-level grammar by combining filler words identified in the creation of the sub-grammars.

13. The system of claim 12, further comprising:

checking means checking the spelling of words in said data.

14. The system of claim 12, wherein said hierarchical grammar is combined with a statistical language model to recognize speech.

15. The system of claim 12, wherein said hierarchical grammar is used to recognize speech.

16. The system of claim 12, wherein said data represents a request to perform said task.

17. The system of claim 12, wherein said data is received using the Internet.

18. The system of claim 12, wherein said parts of speech include verbs and nouns.

19. The system of claim 12, wherein said data is received using an internet-based distributed knowledge capture system.

20. The system of claim 19, wherein said internet-based distributed knowledge capture system is an OpenMind indoor common sense knowledge capture system.

21. The system of claim 12, further comprising:

filler synonym means for identifying synonyms of the identified filler words.

22. The system of claim 12, wherein the high-level grammar is created concurrently with the creation of sub-grammars.