System and method for using a correspondence table to compress a pronunciation guide

• US 5,845,238 A
• Filed: 06/18/1996
• Issued: 12/01/1998
• Est. Priority Date: 06/18/1996
• Status: Expired due to Term

• Current Assignee: Apple Computer Incorporated (Apple Inc.)
• Original Assignee: Apple Computer Incorporated (Apple Inc.)
• Inventors: Fredenburg, Timothy
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): EDOUARD, PATRICK NESTOR
• Application Number:
  US08/665,404
• Time in Patent Office: 896 Days
• Field of Search: 704/1, 704/10, 704/201, 707/530, 707/532, 434/156, 434/169, 341/50, 341/51, 341/55
• US Class Current: 704/1
• CPC Class Codes: G10L 15/187 Phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams
• CPC Class Codes: H03M 7/3084 using adaptive string matching, e.g. the Lempel-Ziv method

Abstract

Parsing routines extract from a conventional pronunciation dictionary an entry, which includes a dictionary word and dictionary phonemes representing the pronunciation of the dictionary word. A correspondence table is used to compress the pronunciation dictionary. The correspondence table includes correspondence sets for a particular language, each set having a correspondence text entry, a correspondence phoneme entry representing the pronunciation of the correspondence text entry and a unique correspondence set identifying symbol. A matching system compares a dictionary entry with the correspondence sets, and replaces the dictionary entry with the symbols representing the best matches. In the absence of a match, symbols representing silent text or unmatched phonemes can be used. The correspondence symbols representing the best matches provide compressed pronunciation dictionary entries. The matching system also generates decoder code sets for subsequently translating the symbol sets. A decoder system uses the decoder code sets for translating symbol sets in the compressed pronunciation dictionary to generate phonemes corresponding to selected text.

Claims

1. A system for compressing a pronunciation guide which includes a plurality of guide entries, each entry having a guide word and at least one associated phoneme representing the pronunciation of the word, the system comprising:

memory storing

(1) a correspondence table which includes a plurality of correspondence sets, each set having

(i) a text entry,

(ii) a phoneme entry representing a pronunciation of the text entry, and

(iii) a symbol identifying the correspondence set; and

(2) a matching system for comparing a selected guide word and the associated phonemes with correspondence sets, and storing correspondence symbols which represent matching correspondence sets as a compressed pronunciation guide entry in the memory; and

a processing unit coupled to the memory for controlling the operations of the matching system.

2. The system of claim 1 wherein the correspondence table includes correspondence sets for productive combinations of phonemes and text in a particular language.

3. The system of claim 1 wherein the correspondence symbols are numbers, each representing the position of the respective correspondence set in the correspondence table.

4. The system of claim 1 wherein the memory further stores a tuning function which enables deletion of unproductive correspondence sets from the correspondence table.

5. The system of claim 1 wherein the matching system compares various correspondence sets, and if several matches are made selects the best match.

6. The system of claim 1 wherein the matching system generates a special symbol representing a silent character, and stores the special symbol as part of a compressed pronunciation guide entry in the memory.

7. The system of claim 1 wherein the matching system generates a special symbol representing a phoneme without any corresponding characters, and stores the special symbol as part of a compressed pronunciation guide entry in the memory.

8. The system of claim 1 wherein the matching system generates a decoder table comprising decoder code sets for use in subsequently de-compressing compressed pronunciation guide entries.

9. The system of claim 8 wherein the decoder code sets replicate a portion of the correspondence table.

10. The system of claim 8 wherein the decoder code sets include symbols representing silent text.

11. The system of claim 8 wherein the decoder code sets include symbols representing phonemes without corresponding characters.

12. The system of claim 1 wherein the matching system selects correspondence sets from the correspondence table for comparison with characters and phonemes from the guide entry.

13. The system of claim 1 wherein the pronunciation guide includes a pronunciation dictionary.

14. A system for using a compressed pronunciation guide and decoder table to decode selected text, comprising:

memory storing

(1) a compressed pronunciation guide having a plurality of symbol sets, each symbol set representing a guide word and at least one corresponding guide phoneme representing the pronunciation of the guide word,

(2) a decoder table having a plurality of decoder code sets for translating symbol sets, each decoder code set including a decoder text entry, a decoder phoneme entry and a decoder symbol representing the decoder code set;

(3) a decoder system for using the decoder table to translate symbol sets to find a guide word which matches the selected text, and upon finding a match using the decoder table

to retrieve the decoder phonemes from the matching symbol set; and

a processor coupled to the memory for controlling the operations of the decoder system.

15. The system of claim 14 wherein the decoder code sets include symbols representing silent text.

16. The system of claim 14 wherein the decoder code sets include symbols representing phonemes without corresponding characters.

17. A computer-based method for compressing a pronunciation guide which includes a plurality of guide entries, each entry having a guide word and at least one associated guide phoneme representing the pronunciation of the guide word, comprising the steps of:

providing a computer memory;

storing in a first portion of the computer memory a correspondence table which includes a plurality of correspondence sets, each correspondence set including a correspondence text entry, a correspondence phoneme entry representing a pronunciation of the correspondence text entry and a unique correspondence symbol identifying the correspondence set;

receiving a guide word and at least one guide phoneme representing the pronunciation of the guide word;

comparing the guide word and guide phonemes with correspondence sets; and

storing the correspondence symbols representing matching correspondence sets as compressed pronunciation guide entries in a second portion of the computer memory.

18. The method of claim 17 wherein the correspondence table includes correspondence sets for productive combinations of phonemes and text in a particular language.

19. The method of claim 17 wherein the correspondence symbol is a number representing the position of the correspondence set in the correspondence table.

20. The method of claim 17 further comprising, after the step of storing in a first portion and before the step of receiving, the step of deleting unproductive correspondence sets from the correspondence table.

21. The method of claim 17 wherein the step of comparing further comprises:

selecting a next currently-unmatched guide phoneme from the guide entry;

retrieving all correspondence sets from the correspondence table which begin with the selected guide phoneme; and

comparing the retrieved correspondence sets with the remaining portions of the guide entry.

22. The method of claim 21 wherein the step of comparing further comprises examining several correspondence sets, and if multiple matches are made selecting the best match.

23. The method of claim 22 wherein the step of comparing further comprises, if a match is not made, generating a special symbol which represents a silent guide character, and storing the special symbol in the memory as part of a compressed pronunciation guide entry.

24. The method of claim 22 wherein the step of comparing further comprises, if a match is not made, generating a special symbol which represents a phoneme without any corresponding guide characters, and storing the special symbol in the memory as part of a compressed pronunciation guide entry.

25. The method of claim 17 and further comprising the step of generating a decoder table including decoder code sets for decompressing the compressed pronunciation guide entries.

26. The method of claim 25 wherein the decoder code sets replicate a portion of the correspondence table.

27. The method of claim 25 wherein the decoder code sets include symbols representing silent text.

28. The method of claim 25 wherein the decoder code sets include symbols representing phonemes without corresponding guide characters.

29. The method of claim 17 further comprising, after the step of storing in a second portion, the step of using the compressed pronunciation guide to generate phonemes representing the selected text.

30. A computer-based method for using a compressed pronunciation guide and a decoder table to retrieve phonemes for selected text, comprising the steps of:

providing computer memory;

storing in a first portion of the computer memory a compressed pronunciation guide which includes a plurality of symbol sets, each symbol set representing a guide word and at least one guide phoneme representing the pronunciation of the guide word;

storing in a second portion of the computer memory a decoder table which includes a plurality of decoder sets, each decoder set having a decoder text entry, a decoder phoneme entry representing the pronunciation of the decoder text entry, and a unique decoder set identifying symbol;

receiving selected text;

using the decoder table to decode a symbol set in the pronunciation guide to produce a guide word;

comparing the selected text with the guide word to determine if they match; and

if a match is made, using the decoder table to retrieve the guide phonemes corresponding to a matching symbol set.

31. A computer storage medium storing a computer program for causing a computer to perform the steps of:

allocating computer memory;

storing in a first portion of the computer memory a compressed pronunciation guide which includes a plurality of symbol sets, each symbol set representing a guide word and guide phonemes representing the pronunciation of the guide word;

storing in a second portion of the computer memory a decoder table which includes a plurality of decoder sets, each decoder set having a decoder text entry, a decoder phoneme entry representing the pronunciation of the decoder text entry, and a unique decoder set identifying symbol;

receiving selected text;

using the decoder table to decode a symbol set in the pronunciation guide to produce a guide word;

comparing the selected text with the guide word to determine if they match; and

if a match is made, using the decoder table to retrieve the guide phonemes corresponding to a matching symbol set.

32. A computer storage medium storing a computer program for causing a computer to perform the steps of:

allocating computer memory;

storing in a first portion of the computer memory a correspondence table which includes a plurality of correspondence sets, each correspondence set including a correspondence text entry, a correspondence phoneme entry representing the pronunciation of the correspondence text entry and a unique correspondence symbol identifying each correspondence set;

receiving a guide word and at least one guide phoneme representing the pronunciation of the guide word;

comparing the guide word and guide phonemes with correspondence sets; and

storing the correspondence symbols representing matching correspondence sets as compressed pronunciation guide entries, in a second portion of the computer memory.

33. A computer-based system for compressing a pronunciation guide, which includes a guide word and at least one guide phoneme representing the pronunciation of the guide word, comprising:

computer memory;

means for storing in a first portion of the computer memory a correspondence table which includes a plurality of correspondence sets, each correspondence set including a correspondence text entry, a correspondence phoneme entry representing the pronunciation of the correspondence text entry, and a unique correspondence symbol identifying the correspondence set;

means for receiving a guide word and at least one guide phoneme representing the pronunciation of the guide word;

means for comparing the guide word and guide phonemes with correspondence sets; and

means for storing the correspondence symbols representing matching correspondence sets as a compressed pronunciation guide entry in a second portion of the computer memory.