Hybrid baseform generation

• US 20040034524A1
• Filed: 08/14/2002
• Published: 02/19/2004
• Est. Priority Date: 08/14/2002
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: International Business Machines Corporation
• Inventors: Rajput, Nitendra, Verma, Ashish
• Granted Patent: US 7,206,738 B2
• Time in Patent Office: Days
• US Class Current: 704/9
• CPC Class Codes: G06F 40/242 Dictionaries

Abstract

A method, a computer system and a computer program product for generating baseforms or phonetic spellings from input text are disclosed. The baseforms are initially generated using rules defined for a particular language. Then, phones are identified in the language that are exceptions to the defined rules and an action is associated with each identified phone. A statistical technique is applied to determine whether the identified phones can be modified. Finally, baseforms containing the identified phones that can be modified, are corrected according to the associated actions. Preferably, the statistical technique is only applied to baseforms containing phones that are exceptions to the defined rules. The defined rules can comprise spelling-to-sound rules for a particular phonetic language that incorporate all possible alternative pronunciations of each baseform.

Claims

1. A method for generating baseforms or phonetic spellings from input text, said method including the steps of:

defining rules for generating a pronunciation dictionary for a particular language;

generating baseforms using said defined rules;

identifying phones, in said language, that are exceptions to said defined rules;

associating an action with each said phone;

applying a statistical technique to determine whether said identified phones can be modified; and

correcting baseforms containing said identified phones that can be modified according to said associated actions.

2. A method according to claim 1, wherein said statistical technique is applied solely to baseforms containing phones that are exceptions to said defined rules.

3. A method according to claim 2, wherein said defined rules comprise spelling-to-sound rules for a particular phonetic language.

4. A method according to claim 3, wherein all possible alternative pronunciations of each baseform are incorporated in said defined rules.

5. A method according to claim 3 or claim 4, wherein said language comprises Hindi.

6. A method according to claim 1 or claim 2, including the further step of applying a statistical technique to reduce a number of alternative baseforms to generate a reduced size pronounciation dictionary.

7. A method according to claim 6, wherein said statistical technique uses information relating to the context of a phone in a baseform.

8. A method according to claim 6, wherein said reduced size pronounciation dictionary is generated by performing one or more steps selected from the group consisting of:

removing a phone from a baseform;

replacing a phone by another phone; and

deleting a baseform.

9. A method according to claim 1 or claim 2, wherein said associated actions comprise one or more actions selected from the group consisting of:

removing a phone from a baseform;

replacing a phone by another phone; and

deleting a baseform.

10. A computer system for generating baseforms or phonetic spellings from input text, including:

processor means for defining rules for generating a pronunciation dictionary for a particular language;

processor means for generating baseforms using said defined rules;

processor means for identifying phones, in said language, that are exceptions to said defined rules;

processor means for associating an action with each said phone;

processor means for applying a statistical technique to determine whether said identified phones can be modified; and

processor means for correcting baseforms containing said identified phones that can be modified according to said associated actions.

11. A computer system according to claim 10, wherein said statistical technique is applied solely to baseforms containing phones that are exceptions to said defined rules.

12. A computer system according to claim 11, wherein said defined rules comprise spelling-to-sound rules for a particular phonetic language.

13. A computer system according to claim 12, wherein all possible alternative pronunciations of each baseform are incorporated in said defined rules.

14. A computer system according to claim 12 or claim 13, wherein said language comprises Hindi.

15. A computer system according to claim 10 or claim 11, further including processor means for applying a statistical technique to reduce a number of alternative baseforms to generate a reduced size pronounciation dictionary.

16. A computer system according to claim 15, wherein said statistical technique uses information relating to the context of a phone in a baseform.

17. A computer system according to claim 15, further including processor means for generating said reduced size pronounciation dictionary, said processor means selected from the group consisting of:

processor means for removing a phone from a baseform;

processor means for replacing a phone by another phone; and

processor means for deleting a baseform.

18. A computer system according to claim 10 or claim 11, wherein said associated actions comprise one or more actions selected from the group consisting of:

removing a phone from a baseform;

replacing a phone by another phone; and

deleting a baseform.

19. A computer program product having a computer readable medium having a computer program recorded therein for generating baseforms or phonetic spellings from input text, said computer program product including:

computer program code means for defining rules for generating a pronunciation dictionary for a particular language;

computer program code means for generating baseforms using said defined rules;

computer program code means for identifying phones, in said language, that are exceptions to said defined rules;

computer program code means for associating an action with each said phone;

computer program code means for applying a statistical technique to determine whether said identified phones can be modified; and

computer program code means for correcting baseforms containing said identified phones that can be modified according to said associated actions.

20. A computer program product according to claim 19, wherein said statistical technique is applied solely to baseforms containing phones that are exceptions to said defined rules.

21. A computer program product according to claim 20, wherein said defined rules comprise spelling-to-sound rules for a particular phonetic language.

22. A computer program product according to claim 21, wherein all possible alternative pronunciations of each baseform are incorporated in said defined rules.

23. A computer program product according to claim 21 or claim 22, wherein said language comprises Hindi.

24. A computer program product according to claim 19 or claim 20, further including computer program code means for applying a statistical technique to reduce a number of alternative baseforms to generate a reduced size pronounciation dictionary.

25. A computer program product according to claim 24, wherein said statistical technique uses information relating to the context of a phone in a baseform.

26. A computer program product according to claim 24, further including processor means for generating said reduced size pronounciation dictionary, said computer program code means selected from the group consisting of:

computer program code means for removing a phone from a baseform;

computer program code means for replacing a phone by another phone; and

computer program code means for deleting a baseform.

27. A computer program product according to claim 19 or claim 20, wherein said associated actions comprise one or more actions selected from the group consisting of:

removing a phone from a baseform;

replacing a phone by another phone; and

deleting a baseform.

28. A computer program product having a computer readable medium having a computer program recorded therein for generating baseforms or phonetic spellings from input text, said computer program product including:

a pronunciation dictionary for a particular language;

a list of actions associated with phones that are exceptions to rules defined for said language;

computer program code means for generating baseforms according to said defined rules;

computer program code means for identifying phones, in said baseforms, that are exceptions to said defined rules;

computer program code means for applying a statistical technique to determine whether said identified phones can be modified; and

computer program code means for correcting baseforms containing said identified phones that can be modified according to said associated actions.

29. A computer program product according to claim 28, wherein said statistical technique is applied solely to baseforms containing phones that are exceptions to said defined rules.

30. A computer program product according to claim 29, wherein said statistical technique uses information relating to the context of a phone in a baseform.

31. A computer program product according to claim 30, wherein said associated actions comprise one or more actions selected from the group consisting of:

removing a phone from a baseform;

replacing a phone by another phone; and

deleting a baseform.

32. A computer program product according to claim 31, wherein said defined rules comprise spelling-to-sound rules for a particular phonetic language.

33. A computer program product according to any one of claims 28 to 32, wherein said language comprises Hindi.