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

• US 6,178,397 B1
• Filed: 10/08/1998
• Issued: 01/23/2001
• 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): Chang, Vivian
• Assistant Examiner(s): EDOUARD, PATRICK NESTOR
• Application Number:
  US09/169,190
• Time in Patent Office: 838 Days
• Field of Search: 704/1, 704/9, 704/10, 704/258, 704/260, 707/532, 707/101
• 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 computer data storage medium storing a correspondence table which enables compression of a pronunciation dictionary, the correspondence table comprising:

a plurality of correspondence sets each including

a correspondence text entry that is part of a dictionary word;

a correspondence phoneme entry representing the pronunciation of the correspondence text entry; and

a correspondence symbol for identifying the correspondence set,

wherein at least one said correspondence symbol forms a symbol set for use as a compressed data entry in generating said compressed pronunciation dictionary.

2. The computer data storage medium of claim 1 further storing a tuning function for optimizing said correspondence table.

3. The computer data storage medium of claim 2 wherein said tuning function eliminates redundant correspondence sets and low usage correspondence sets from said correspondence table.

4. The computer data storage medium of claim 1 wherein said correspondence table includes said correspondence sets for all practical combinations of said correspondence text entries and said correspondence phoneme entries for a given language.

5. The computer data storage medium of claim 1 further storing:

a grouping of a plurality of said correspondence sets.

6. The computer data storage medium of claim 5 wherein correspondence phoneme entries of said grouping are similar to one another in pronunciation.

7. The method of claim 1 wherein a matching system uses said correspondence phoneme entry to match said correspondence sets in generating said compressed pronunciation dictionary.

8. A system for storing a pronunciation guide comprising:

a correspondence table for storing pronunciation data; and

a tuning function for optimizing said correspondence table;

wherein said correspondence table includes at least one correspondence set having

a correspondence text entry that is part of a dictionary word,

a correspondence phonetic entry representing the pronunciation of said correspondence text entry, and

a correspondence symbol for identifying the correspondence set; and

wherein a matching system uses said correspondence phonetic entry to match said at least one correspondence set in generating a compressed pronunciation dictionary.

9. The system of claim 8 wherein said tuning function eliminates redundant correspondence sets from said correspondence table.

10. The system of claim 9 wherein at least one said correspondence symbol forms a symbol set for use as a compressed entry in generating said compressed pronunciation dictionary.

11. The system of claim 9 wherein said correspondence table includes a respective correspondence set for each practical combination of any correspondence text entry and any correspondence phonetic entry for a given language.

12. The system of claim 9 further comprising:

a grouping of a plurality of said at least one correspondence set.

13. The system of claim 12 wherein said phonetic entries of said grouping are similar to one another in pronunciation.

14. The system of claim 8 wherein said tuning function eliminates low usage correspondence sets from said correspondence table.

15. The system of claim 8 wherein said phonetic entry is a phoneme, an allophone, or a syllable.

16. A method of storing a pronunciation guide, comprising the steps of:

inputting a correspondence set into a correspondence table; and

inputting into said correspondence table a correspondence symbol corresponding to said correspondence set;

wherein at least one said correspondence symbol forms a symbol set for use as a compressed entry in generating a compressed pronunciation dictionary.

17. The method of claim 16 further comprising the steps of:

optimizing said correspondence table; and

grouping said correspondence set into a plurality of said correspondence sets.

18. The method of claim 17 wherein said step of optimizing comprises the steps of:

eliminating redundant correspondence sets from said correspondence table;

eliminating low-usage correspondence sets from said correspondence table; and

adding productive correspondence sets to said correspondence table.

19. The method of claim 16 wherein said step of inputting a correspondence set further comprises the steps of:

inputting a correspondence text entry that is part of a dictionary word into said correspondence table; and

inputting a phonetic entry corresponding to said correspondence text entry into said correspondence table.

20. The method of claim 19 further comprising the step of using said phonetic entry to generate said compressed pronunciation dictionary.