Automated grammar generator (AGG)

• US 20050154580A1
• Filed: 10/28/2004
• Published: 07/14/2005
• Est. Priority Date: 10/30/2003
• Status: Abandoned Application

• Current Assignee: Vox Generation Ltd.
• Original Assignee: Vox Generation Ltd.
• Inventors: Buckley, Pierce, Horowitz, David
• Application Number:
  US10/976,030
• Publication Number:
  US 20050154580A1
• Time in Patent Office: Days
• US Class Current: 704/9
• CPC Class Codes: G06F 40/211 Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
• CPC Class Codes: G06F 40/268 Morphological analysis
• CPC Class Codes: G06F 40/289 Phrasal analysis, e.g. finite state techniques or chunking
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: G10L 15/19 Grammatical context, e.g. disambiguation of the recognition hypotheses based on word sequence rules

Abstract

An automated grammar generator is disclosed, which is operable to receive a speech or text segment. The automated grammar generator identifies one or more parts of the segment suitable for processing into a natural language expression. The natural language expression is an expression which a person might use to refer to the segment. The automatic grammar generator generates one or more phrases from the segment, each of the one or more phrases corresponding to or capable of it being processed into a natural language expression or utterance suitable for referencing the text or speech segment. Noun phrases and verb phrases and other syntactic structures are identified in the speech or text segment, and modified to produce typical natural language expressions or utterances a user might employ to reference a segment. Verbs in verb phrases may be modified in order to provide further natural language expressions or utterances for use in the grammar. The natural language expressions thus generated may be included in grammars or language models to produce models for recognition using an automatic speech recogniser in a spoken language interface.

Claims

1. An automated grammar generator, operable to:

receive a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

3. An automated grammar generator according to claim 1, comprising a phrase chunking module operable to generate automatically at least one phrase from said at least one part of said segment, said at least one phrase corresponding to at least one natural language expression.

4. An automated grammar generator according to claim 3, further comprising a term extraction module operable to identify a syntactic phrase in said segment;

wherein said phrase chunking module is operable to generate at least one variation of said syntactic phrase, thereby automatically generating said at least one phrase.

5. An automated grammar generator according to claim 4, wherein:

said term extraction module is operable to identify a noun phrase in said segment; and

said phrase chunking module is operable to generate at least one phrase comprising at least one noun from said noun phrase.

6. An automated grammar generator according to claim 5, wherein said term extraction module is operable to identify in said segment a noun phrase comprising a plurality of nouns.

7. An automated grammar generator according to claim 4, wherein said term extraction module is further operable to include within a general class of noun the following parts of speech:

proper noun, singular or mass noun, plural noun, adjective, cardinal number, and adjective superlative.

8. An automated grammar generator according to claim 5, wherein said phrase chunking module is further operable to associate at least one adjective with said noun phrase in at least one of said at least one phrase.

9. An automated grammar generator according to of claim 3, wherein:

said term extraction module is operable to identify a verb phrase in said segment; and

said phrase chunking module is operable to generate at least one phrase comprising at least one verb from said verb phrase.

10. An automated grammar generator according to claim 9, wherein said phrase chunking module is further operable to associate at least one adverbs with said verb phrase in at least one of said at least one phrase.

11. An automated grammar generator according to claim 9, further comprising a morphological variation module operable to modify a tense of said verb phrase to generate said at least one phrase.

12. An automated grammar generator according to claim 11, wherein said morphological variation module is operable to identify the stem of a verb in said verb phrase and add an ending to said stem to modify said tense.

13. An automated grammar generator according to claim 11, wherein said morphological variation module is operable to vary the constituents of said verb phrase to modify said tense.

14. An automated grammar generator according to claim 11, wherein said morphological variation module is operable to add the word “

being”

before the past tense of a verb in said verb phase.

2. An automated grammar generator, operable to:

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

48. An automated grammar generator according to claim 2, further comprising a phrase chunking module operable to generate automatically at least one phrase from said at least one part of said segment, said at least one phrases corresponding to at least one natural language expressions.

15. An automated speech recognition system comprising an automated grammar generator operable to:

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

16. A spoken language interface comprising an automated grammar generator operable to;

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

17. A spoken language interface according to claim 16, further operable to support a multi-modal input and/or output environment thereby to provide output and/or receive input information on at least one of the following modalities:

keyed, text spoken, audio, written, and graphic.

18. A computer system comprising an automated grammar generator operable to:

receive a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

19. An automated information service comprising:

a spoken language interface, wherein the spoken language interface comprises an automated grammar generator operable to;

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

20. An automated information service according to claim 19 comprising at least one of the following services:

a news service;

a sports report service;

a travel information service;

an entertainment information service;

an e-mail response system;

an internet search engine interface;

an entertainment service;

a cinema ticket booking;

catalogue searching;

TV programme listings;

navigation service;

equity trading service;

warehousing and stock control;

distribution queries;

Customer Relationship Management;

medical service/patient records; and

interfacing to a hospital data.

21. A user device comprising an automated grammar generator operable to:

receive a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

22. A communications system comprising:

a computer system comprising an automated grammar generator

operable to;

receive a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment;

and a user device,

wherein said computer system and said user device are operable to communicate with each other over a communications network, and

wherein said user device is operable to transmit one of a text segment and a speech segment to said computer system over said communications network, for said computer system generating a grammar for referencing said segment.

23. A method of operating a computer system for automatically generating a grammar comprising:

receiving a text segment; and

identifying at least one part of the text segment suitable for processing into a natural language expression for referencing the segment, said natural language expression being an expression a human might use to refer to said segment.

25. The method of claim 23, further comprising automatically generating at least one phrase from said at least one part of said segment wherein said at least one phrase correspond to at least one natural language expression.

26. The method of claim 25, further comprising identifying a syntactic phrase of said segment and generating at least one variation of said syntactic phrase, thereby automatically generating said at least one phrase.

27. The method of claim 26, further comprising:

identifying a noun phrase of said segment; and

generating at least one phrase comprising at least one noun from said noun phrase.

28. The method of claim 27, further comprising identifying a noun phrase comprising more than one noun in said segment.

29. The method of claim 27, further comprising including one or more adjectives associated with said noun phrase in at least one of said at least one phrase.

30. The method of claim 27, further comprising clarifying within a general class of noun the following parts of speech:

proper noun, singular of mass noun, plural noun, adjective, cardinal number, and adjective superlative.

31. The method of claim 23, further comprising.

identifying a verb phrase in said segment;

and

generating one or more phrases comprising one or more verbs from said verb phrase.

32. The method of claim 31, further comprising including at least one adverb associated with said verb phrase in at least one of said at least one phase.

33. The method of claim 31, further comprising automatically modifying a tense of said verb phrase to generate said at least one phrase.

34. The method of claim 31, further comprising identifying the stem of a verb in said verb phrase and adding an ending to said stem to modify said tense.

35. The method of claim 31, further comprising varying the constituents of said verb phrase to modify said tense.

36. The method of claim 34, further comprising adding the word “

being”

before the past tense of a verb phrase.

24. A method of operating a computer system for automatically generating a grammar comprising:

receiving a speech segment;

converting said speech segment into a text segment; and

identifying at least one part of the text segment suitable for processing into a natural language expression for referencing the segment, said natural language expression being an expression a human might use to refer to said segment.

49. The method of claim 24, further comprising automatically generating at least one phrase from said at least one part of said segment wherein said at least one phrase correspond to at least one natural language expression.

37. A computer program for implementing an automated grammar generator, the automated grammar generator operable to:

receive a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

38. A computer usable carrier medium carrying a computer program for implementing an automated grammar generator, the automated grammar generator operable to:

receive a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

39. (canceled)

40. (canceled)

41. (canceled)

42. (canceled)

43. (canceled)

44. An automated grammar generator, comprising:

means for receiving a text segment; and

means for identifying at least one part of said text segment for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

45. An automated grammar generator, comprising:

means for receiving a speech segment;

means for converting said speech segment into a text segment; and

means for identifying at least one part of said text segment for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

46. A method of operating a computer system for automatically generating a grammar, comprising:

a step for receiving a text segment; and

a step for identifying at least one part of said text segment for processing into a natural language expression fore referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

47. A method of operating a computer system for automatically generating a grammar, comprising:

a step for receiving a speech segment;

a step for converting said speech segment into a text segment; and

a step for identifying at least one part of said segment for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

50. A computer system comprising an automated grammar generator configured to:

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

51. A computer system comprising an automated speech recognition system, the automated speech recognition system comprising an automated grammar generator operable to:

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

52. A computer system comprising a spoken language interface, the spoken language interface comprising an automated grammar generator operable to:

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

53. A user device comprising an automated grammar generator configured to:

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

54. A user device comprising an automated speech recognition system, the automated speech recognition system comprising an automated grammar generator operable to:

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

55. A user device comprising spoken language interface, the spoken language interface comprising an automated grammar generator operable to:

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

56. A computer program for implementing an automated grammar generator configured to:

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

57. A computer program for implementing an automated speech recognition system, the automated speech recognition system comprising an automated grammar generator operable to:

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

58. A computer program for implementing a spoken language interface, the spoken language interface comprising an automated grammar generator operable to:

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

59. A computer program for operating a computer system, comprising an automated grammar generator operable to:

receive a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

60. A computer program for operating a computer system, comprising an automated grammar generator operable to:

receive a speech segment;

convert said speech segment into a text segment; and

identify at least one part of said text segment suitable for processing into a natural language expression for referencing said segment, said natural language expression being an expression a human might use to refer to said segment.

61. A computer program for implementing a method of operating a computer system for automatically generating a grammar comprising:

receiving a text segment; and

identifying at least one part of the text segment suitable for processing into a natural language expression for referencing the segment, said natural language expression being an expression a human might use to refer to said segment.

62. A computer program for implementing a method of operating a computer system for automatically generating a grammar comprising:

receiving a speech segment;

converting said speech segment into a text segment; and

identifying at least one part of the text segment suitable for processing into a natural language expression for referencing the segment, said natural language expression being an expression a human might use to refer to said segment.