Language independent parsing in natural language systems

• US 20070185702A1
• Filed: 02/09/2006
• Published: 08/09/2007
• Est. Priority Date: 02/09/2006
• Status: Active Grant

• Current Assignee: Janet Dwyer, John Harney
• Original Assignee: Janet Dwyer, John Harney
• Inventors: Harney, John, Dwyer, Janet
• Granted Patent: US 8,229,733 B2
• Time in Patent Office: Days
• US Class Current: 704/4
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: G06F 40/211 Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
• CPC Class Codes: G06F 40/242 Dictionaries

Abstract

There are provided methods and apparatus for parsing and for self-learning in natural language systems. A parsing method for a natural language system includes the step of parsing an input phrase to identify at least one word or phrase in the input phrase for which synonyms exist. The method further includes the step of substituting the synonyms in place of the identified at least one word or phrase, in descending order by text length, to provide a modified input phrase. The method also includes the step of repeating the parsing and substituting steps one of until the synonyms no longer exist or a pre-specified number of times.

Claims

1. A parsing method for a natural language system, comprising:

parsing an input phrase to identify at least one word or phrase in the input phrase for which synonyms exist;

substituting the synonyms in place of the identified at least one word or phrase, in descending order by text length, to provide a modified input phrase; and

repeating said parsing and substituting steps until the synonyms no longer exist or a pre-specified number of times.

2. The method of claim 1, wherein the text length corresponds to at least one of the synonyms and the identified at least one word or phrase.

3. The method of claim 1, wherein the input phrase is in a non-formal language.

4. The method of claim 1, further comprising maintaining a mapping of at least one of words-to-synonyms and phrases-to-synonyms.

5. The method of claim 4, wherein the input phrase and at least some words and phrases in the mapping are in a non-formal language.

6. The method of claim 1, further comprising:

parsing the modified input phrase to identify commands therein, subsequent to said repeating step; and

replacing the identified commands with language tags associated with the identified commands, wherein the associated language tags correspond to a target application program.

7. The method of claim 6, wherein said replacing step replaces the identified commands with the associated language tags in ascending order by text length.

8. The method of claim 6, wherein the text length corresponds to at least one of the language tags and the identified commands.

9. The method of claim 6, further comprising maintaining a mapping of commands-to-associated-language-tags.

10. The method of claim 6, wherein the language tags are in a computer markup language.

11. A parsing method for a natural language system, comprising:

parsing an input phrase to identify commands therein; and

replacing the commands with language tags associated with the commands, in ascending order by text length,

wherein the associated language tags correspond to a predefined library of functions in a target application program.

12. The method of claim 11, wherein the text length corresponds to at least one of the language tags and the identified commands.

13. The method of claim 11, further comprising maintaining a mapping of commands-to-associated-language-tags.

14. The method of claim 11, wherein the language tags are in a computer markup language.

15. The method of claim 11, wherein said step of parsing the modified input phrase comprises:

detecting a new command indicator in the input phrase, the new command indicator configured to indicate a presence of a new user-defined command in the input phrase and to facilitate parsing of the new user-defined command from the input phrase; and

utilizing the new command indicator to extract the new user-defined command from the input phrase for further processing.

16. The method of claim 15, wherein the new command indicator facilitates parsing of the new user-defined command from the input phrase by specifying at least one of a start boundary and an end boundary of the new user-defined command within the input phrase.

17. The method of claim 15, wherein the method further comprises creating a mapping for the new user-defined command with respect to at least one other existing command also present in the input phrase, using a built-in mapping format of the new command indicator that maps the new user-defined command to the at least one other existing command within the input phrase.

18. The method of claim 15, wherein the new command indicator includes a mapping format configured to map the new user-defined command to more than one existing command to form a command set, and the input phrase is parsed to extract the new user-defined command and the more than one existing command from the input phrase using the mapping format.

19. The method of claim 15, wherein the new user-defined command is in a non-formal language.

20. An apparatus for parsing an input phrase in a natural language system, comprising:

a synonym mapper configured to map at least one of words-to-synonyms and phrases-to-synonyms; and

a parser, in signal communication with said synonym mapper, configured to parse the input phrase to identify at least one word or phrase in the input phrase for which synonyms exist, substitute the synonyms in place of the identified at least one word or phrase in descending order by text length to provide a modified input phrase, and repeat the parsing and the substituting until the synonyms no longer exist or a pre-specified number of times.

21. The apparatus of claim 20, wherein the text length corresponds to at least one of the synonyms and the identified at least one word or phrase.

22. The apparatus of claim 20, wherein the input phrase is in a non-formal language.

23. The method of claim 20, wherein at least some words and phrases mapped by said synonym mapper are in the non-formal language.

24. The apparatus of claim 20, wherein the apparatus is implemented as a language independent dynamic grammar engine with self-learning capability.

25. The apparatus of claim 20, wherein said parser is further configured to parse the modified input phrase to identify commands therein, and to replace the identified commands with language tags associated with the identified commands, wherein the associated language tags correspond to a predefined command library in target application program.

26. The apparatus of claim 25, wherein said parser is configured to replace the identified commands with the associated language tags in ascending order by text length.

27. The apparatus of claim 26, wherein the text length corresponds to at least one of the language tags and the identified commands.

28. The apparatus of claim 25, further comprising a language tag mapper for mapping commands-to-associated-language-tags.

29. The apparatus of claim 25, wherein the language tags are in a computer markup language.

30. An apparatus for parsing an input phrase in a natural language system, comprising:

a language tag mapper configured to map commands-to-associated-language-tags; and

a parser, in signal communication with said language tag mapper, configured to parse an input phrase to identify commands therein, and to replace the commands with language tags associated with the commands in ascending order by text length,

wherein the associated language tags correspond to a target application program.

31. The apparatus of claim 30, wherein the text length corresponds to at least one of the language tags and the identified commands.

32. The apparatus of claim 30, wherein the language tags are in a computer markup language.

33. The apparatus of claim 30, wherein said parser is configured to detect a new command indicator in the input phrase, the new command indicator for indicating a presence of a new user-defined command in the input phrase and for facilitating parsing of the new user-defined command from the input phrase, said parser being further configured to utilize the new command indicator to extract the new user-defined command from the input phrase for further processing.

34. The apparatus of claim 33, wherein the new command indicator facilitates parsing of the new user-defined command from the input phrase by specifying at least one of a start boundary and an end boundary of the new user-defined command within the input phrase.

35. The apparatus of claim 33, wherein said language tag mapper is configured to create a mapping for the new user-defined command with respect to at least one other existing command also present in the input phrase, using a built-in mapping format of the new command indicator that maps the new user-defined command to the at least one other existing command within the input phrase.

36. The apparatus of claim 33, wherein the new command indicator includes a mapping format configured to map the new user-defined command to more than one existing command to form a command set, and said parser is configured to parse the input phrase to extract the new user-defined command and the more than one existing command from the input phrase using the mapping format.

37. The apparatus of claim 33, wherein the new user-defined command is in a non-formal language.

38. A self-learning method for a natural language system, comprising:

detecting a new command indicator in an input phrase, the new command indicator configured to indicate a presence of a new user-defined command in the input phrase and to facilitate parsing of the new user-defined command from the input phrase; and

utilizing the new command indicator to extract the new user-defined command from the input phrase for further processing.

39. The method of claim 38, wherein the new command indicator facilitates parsing of the new user-defined command from the input phrase by specifying at least one of a start boundary and an end boundary of the new user-defined command within the input phrase.

40. The method of claim 38, wherein the method further comprises creating a mapping for the new user-defined command with respect to at least one other existing command also present in the input phrase, using a built-in mapping format of the new command indicator that maps the new user-defined command to the at least one other existing command within the input phrase.

41. The method of claim 38, wherein the new command indicator includes a mapping format configured to map the new user-defined command to more than one existing command to form a command set, and the input phrase is parsed to extract the new user-defined command and the more than one existing command from the input phrase using the mapping format.

42. The method of claim 38, further comprising maintaining a mapping of commands-to-associated-language-tags, wherein the associated language tags correspond to a target application program.

43. The method of claim 38, wherein the new user-defined command is in a non-formal language.

44. A self-learning apparatus for a natural language system, comprising:

a language tag mapper for mapping commands-to-associated-language-tags; and

a parser, in signal communication with said language tag mapper, configured to detect a new command indicator in the input phrase, the new command indicator for indicating a presence of a new user-defined command in the input phrase and for facilitating parsing of the new user-defined command from the input phrase, said parser being further configured to utilize the new command indicator to extract the new user-defined command from the input phrase for further processing.

45. The apparatus of claim 44, wherein the new command indicator facilitates parsing of the new user-defined command from the input phrase by specifying at least one of a start boundary and an end boundary of the new user-defined command within the input phrase.

46. The apparatus of claim 44, wherein said language tag mapper is configured to create a mapping for the new user-defined command with respect to at least one other existing command also present in the input phrase, using a built-in mapping format of the new command indicator that maps the new user-defined command to the at least one other existing command within the input phrase.

47. The method of claim 44, wherein the new command indicator includes a mapping format configured to map the new user-defined command to more than one existing command to form a command set, and said parser is configured to parse the input phrase to extract the new user-defined command and the more than one existing command from the input phrase using the mapping format.

48. The apparatus of claim 44, wherein the new user-defined command is in a non-formal language.