Construction of phonetic representation of a string of characters

• US 9,910,836 B2
• Filed: 12/21/2015
• Issued: 03/06/2018
• Est. Priority Date: 12/21/2015
• Status: Active Grant

• Current Assignee: VeriSign, Inc.
• Original Assignee: VeriSign, Inc.
• Inventors: Raemy, Vincent, Russo, Vincenzo, Hennebert, Jean, Wicht, Baptiste
• Primary Examiner(s): SPOONER, LAMONT M
• Application Number:
  US14/976,968
• Publication Number:
  US 20170177550A1
• Time in Patent Office: 806 Days
• Field of Search: 704 1, 704 9, 704 10, 704257
• CPC Class Codes: G06F 40/157 using dictionaries or tables
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: H04L 61/302 Administrative registration, e.g. for domain names at internet corporation for assigned names and numbers [ICANN]
• CPC Class Codes: H04L 61/3025 Domain name generation or assignment

Abstract

Provided are methods, devices, and computer-readable media for accessing a string of characters; parsing the string of characters into string of graphemes; determining one or more phonetic representations for one or more graphemes in the string of graphemes based on a first data structure; determining at least one grapheme representation for one or more of the one or more phonetic representations based on a second data structure; and constructing the phonetic representation of the string of characters based on the grapheme representation that was determined.

Claims

1. A method, comprising:

accessing a string of characters;

parsing the string of characters into a first string of graphemes;

adding one or more characters to the first string of graphemes to represent missing characters in the string of characters to create a second string of graphemes;

grouping the second string of graphemes into a plurality of pseudo-graphemes based on a probability determined by a trained discrete estimator, wherein two or more graphemes in the string of graphemes that are phonetized together are grouped to a single pseudo-grapheme;

accessing a first data structure that maps each pseudo-grapheme in the string of pseudo-graphemes to one or more universal phonetic representations based on an international phonetic alphabet, wherein the first data structure comprises a plurality of first nodes with each first node of the plurality of first nodes having a respective weight assigned that corresponds to a pronunciation of a grapheme;

determining one or more phonetic representations for each pseudo-grapheme in the string of pseudo-graphemes based on the first data structure;

accessing a second data structure that maps the one or more universal phonetic representations to one or more graphemes, wherein the second data structure comprises a plurality of second nodes with each second node of the plurality of second nodes having a respective weight assigned that corresponds to a likely representation of a grapheme;

determining at least one grapheme representation for one or more of the one or more phonetic representations based on the second data structure; and

constructing a phonetic representation of the string of characters based on the at least one grapheme representation that was determined.

2. The method of claim 1, further comprising:

ranking each grapheme representation to produce a ranked list, wherein the ranking is based on a likelihood that a grapheme representation sounds similar to a pronunciation sound of the string of characters; and

filtering the ranked list to produce a subset of grapheme representations.

3. The method of claim 2, further comprising determining a first composite weight for the one or more phonetic representations based on the first data structure.

4. The method of claim 2, further comprising determining a second composite weight for the one or more graphemes based on the second data structure.

5. The method of claim 4, wherein the filtering is based on the second composite weight.

6. The method of claim 1, further comprising creating the first data structure and the second data structure as information gain trees.

7. A device, comprising:

a memory storing instructions; and

at least one processor, operably connected to the memory, implemented at least in part in hardware, and configured to execute the instructions to perform operations comprising;

receiving the string of characters;

parsing the string of characters into a first string of graphemes;

adding one or more characters to the first string of graphemes to represent missing characters in the string of characters to create a second string of graphemes;

grouping the second string of graphemes into a plurality of pseudo-graphemes based on a probability determined by a trained discrete estimator, wherein two or more graphemes in the string of graphemes that are phonetized together are grouped to a single pseudo-grapheme;

accessing a first data structure that maps each pseudo-grapheme in the string of pseudo-graphemes to one or more universal phonetic representations based on an international phonetic alphabet, wherein the first data structure comprises a plurality of first nodes with each first node of the plurality of first nodes having a respective weight assigned that corresponds to a likely pronunciation of a grapheme;

determining one or more phonetic representations for each pseudo-grapheme in the string of pseudo-graphemes based on the first data structure;

accessing a second data structure that maps the one or more universal phonetic representations to one or more graphemes, wherein the second data structure comprises a plurality of second nodes with each second node of the plurality of second nodes having a respective weight assigned that corresponds to a likely representation of a grapheme;

determining at least one grapheme representation for one or more of the one or more phonetic representation based on the second data structure; and

constructing a phonetic representation of the string of characters based on the at least one grapheme representation that was determined.

8. The device of claim 7, the operations further comprising:

ranking each grapheme representation to produce a ranked list, wherein the ranking is based on a likelihood that a grapheme representation sounds similar to a pronunciation sound of the string of characters; and

filtering the ranked list to produce a subset of grapheme representations.

9. The device of claim 7, the operations further comprising creating the first data structure and the second data structure as information gain trees.

10. The device of claim 7, the operations further comprising determining a first composite weight for the one or more phonetic representations based on the first data structure.

11. The device of claim 7, further comprising determining a second composite weight for the one or more graphemes based on the second data structure.

12. The device of claim 11, wherein the filtering is based on the second composite weight.

13. A non-transitory computer-readable medium comprising computer-interpretable instructions which, when executed by at least one electronic processor, cause the at least one electronic processor to perform a method of converting a string of characters into a phonetic representation, the method comprising:

receiving the string of characters;

parsing the string of characters into a first string of graphemes;

adding one or more characters to the first string of graphemes to represent missing characters in the string of characters to create a second string of graphemes;

grouping the second string of graphemes into a plurality of pseudo-graphemes based on a probability determined by a trained discrete estimator, wherein two or more graphemes in the string of graphemes that are phonetized together are grouped to a single pseudo-grapheme;

accessing a first data structure that maps graphemes to one or more universal phonetic representations based on an international phonetic alphabet, wherein the first data structure comprises a plurality of first nodes with each first node of the plurality of first nodes having a respective weight assigned that corresponds to a likely pronunciation of a grapheme;

determining one or more phonetic representations for each pseudo-grapheme in the string of pseudo-graphemes based on the first data structure;

accessing a second data structure that maps the one or more universal phonetic representations to one or more graphemes, wherein the second data structure comprises a plurality of second nodes with each second node of the plurality of second nodes having a respective weight assigned that corresponds to a likely representation of a grapheme;

determining at least one grapheme representation for one or more of the one or more phonetic representation based on the second data structure; and

constructing the phonetic representation of the string of characters based on the at least one grapheme representation that was determined.

14. The non-transitory computer-readable medium of claim 13, the method further comprising:

ranking each grapheme representation to produce a ranked list, wherein the ranking is based on a likelihood that a grapheme representation sounds similar to a pronunciation sound of the string of characters; and

filtering the ranked list to produce a subset of grapheme representations.

15. The non-transitory computer-readable medium of claim 13, the method further comprising creating the first data structure and the second data structure as information gain trees.

16. The non-transitory computer-readable medium of claim 13, the method further comprising determining a first composite weight for the one or more phonetic representations based on the first data structure.

17. The non-transitory computer-readable medium of claim 13, the method further comprising determining a second composite weight for the one or more graphemes based on the second data structure.

18. The non-transitory computer-readable medium of claim 17, wherein the filtering is based on the second composite weight.