Voice-to-text data processing

1. A method comprising:

converting, based on rules, a word spoken by a user into a pattern of symbols in response to an unsuccessful attempt to retrieve the word in a list, with the pattern of symbols providing a visual representation of speech sounds identifying the word in the list;

comparing the pattern of symbols of the converted word to a database of patterns, with the patterns in the database being in a format of symbols corresponding to the words in the list, and with each pattern used in the compare having a match value assigned thereto based on being compared to the pattern of symbols of the converted word, wherein comparing the pattern of symbols comprises:

dividing the pattern of symbols into pronunciation sections for the converted word, each pronunciation section corresponding to a syllable in the converted word;

dividing the pattern of symbols into pronunciation sections for each pattern in the database used in the compare; and

comparing the pronunciation sections for the converted word to the corresponding pronunciation sections for each pattern used in the compare; and

providing the word in the list corresponding to the pattern having the match value that is indicative of a match to the converted word.

2. The method according to claim 1 , wherein the word spoken by the user is converted into alphabet characters before converting into the pattern of pronunciation symbols, and wherein the unsuccessful attempt to retrieve the word in the list is based on there not being a match between the alphabet characters representing the spoken word to alphabet characters representing the word in the list.

3. The method according to claim 1 , further comprising:

ranking the match values assigned to the patterns used in the compare; and

selecting the pattern having a highest ranked match value that exceeds a threshold.

4. The method according to claim 1 , wherein the patterns in the database comprise a plurality of multi-lingual patterns, with each language pattern in the multi-lingual patterns being based on a particular language pronunciation of the word in the list, and with each word in the list being represented by more than one language pattern.

5. The method according to claim 4 , wherein the comparing starts with a first one of language patterns having a particular language pronunciation, and in response to there not being a match, repeats the comparing with a second one of the language patterns having a different particular language pronunciation.

6. The method according to claim 4 , wherein in response to there being a match with one of the language patterns having a particular language pronunciation, further comprising:

adding the language pattern providing the match to a custom pattern section in the database;

for a next time a word is spoken by the user for retrieval, starting the compare using the language pattern in the custom pattern section; and

in response to there not being a match with the language pattern in the custom pattern section, continuing the compare with the other language patterns in the database.

7. The method according to claim 1 , wherein the matching value assigned to each pattern used in the compare is based on a respective similarity value assigned to each pronunciation section.

8. The method according to claim 7 , wherein the matching value assigned to each pattern is determined by adding the respective similarity values assigned to the pronunciation sections for the pattern, and dividing the added respective similarity values by a number of the pronunciation sections in the converted word.

9. The method according to claim 7 , wherein each pronunciation section comprises a plurality of letters, and wherein the respective similarity value assigned to each pronunciation section for the pattern used in the compare is based on the following:

determining a similarity value for each letter in the pronunciation section for the pattern used in the compare;

multiplying the similarity value for each letter in the pronunciation section by a respective weighting factor; and

adding together the determined similarity value for each letter multiplied by the respective weighting factor to determine the similarity value assigned to each pronunciation section.

10. The method according to claim 1 , wherein the rules for converting the word into a pattern of pronunciation symbols are based on an international phonetic alphabet (IPA).

11. The method according to claim 1 , wherein retrieval of the word spoken is initiated by the user in response to the user speaking a predetermined word.

12. A computing device comprising:

a memory and a processor configured to cooperate with said memory to perform the following:

convert, based on rules, a word spoken by a user into a pattern of pronunciation symbols in response to an unsuccessful attempt to retrieve the word in a list, with the pattern of pronunciation symbols providing a visual representation of speech sounds identifying the word in the list;

compare the pattern of pronunciation symbols of the converted word to a database of patterns, with the patterns in the database being in a format of pronunciation symbols corresponding to the words in the list, and with each pattern used in the compare having a match value assigned thereto based on being compared to the pattern of pronunciation symbols of the converted word, by:

dividing the pattern of symbols into pronunciation sections for the converted word, each pronunciation section corresponding to a syllable in the converted word;

dividing the pattern of symbols into pronunciation sections for each pattern in the database used in the compare; and

comparing the pronunciation sections for the converted word to the corresponding pronunciation sections for each pattern used in the compare; and

provide the word in the list corresponding to the pattern having the match value that is indicative of a match to the converted word.

13. The computing device according to claim 12 , wherein said processor is configured to convert a contact spoken by the user into alphabet characters before converting into the pattern of pronunciation symbols, and wherein the unsuccessful attempt to retrieve the contact in a contact list is based on there not being a match between the alphabet characters representing the spoken contact to the alphabet characters representing the contacts in the contact list.

14. The computing device according to claim 12 , wherein said processor is configured to perform the following:

rank the match values assigned to the patterns used in the compare; and

select the pattern having a highest ranked match value that exceeds a threshold.

15. The computing device according to claim 12 , wherein the patterns in the database comprise a plurality of multi-lingual patterns, with each language pattern in the multi-lingual patterns being based on a particular language pronunciation of a contact in a contact list, and with each contact in the contact list being represented by more than one language pattern.

16. The computing device according to claim 12 , wherein the matching value assigned to each pattern used in the compare is based on a respective similarity value assigned to each pronunciation section.

17. The computing device according to claim 12 , wherein the rules for converting a contact into a pattern of pronunciation symbols are based on an international phonetic alphabet (IPA).

18. The computing device according to claim 12 , wherein each pronunciation section comprises a plurality of letters, and wherein the processor is configured to determine the respective similarity value assigned to each pronunciation section for the pattern used in the compare by:

determining a similarity value for each letter in the pronunciation section for the pattern used in the compare;

multiplying the similarity value for each letter in the pronunciation section by a respective weighting factor; and

adding together the determined similarity value for each letter multiplied by the respective weighting factor to determine the similarity value assigned to each pronunciation section.

19. The computing device according to claim 15 , wherein the processor is further configured to perform:

adding the language pattern providing the match to a custom pattern section in the database;

for a next time a word is spoken by the user for retrieval, starting the compare using the language pattern in the custom pattern section; and

in response to there not being a match with the language pattern in the custom pattern section, continuing the compare with the other language patterns in the database.

20. A non-transitory computer readable medium for a computing device, and with the non-transitory computer readable medium having a plurality of computer executable instructions for causing the computing device to perform steps comprising:

converting, based on rules, a word spoken by a user into a pattern of pronunciation symbols in response to an unsuccessful attempt to retrieve the word in a list, with the pattern of pronunciation symbols providing a visual representation of speech sounds identifying the word in the list;

comparing the pattern of pronunciation symbols of the converted word to a database of patterns, with the patterns in the database being in a format of pronunciation symbols corresponding to the words in the list, and with each pattern used in the compare having a match value assigned thereto based on being compared to the pattern of pronunciation symbols of the converted word, wherein comparing the pattern of symbols comprises:

dividing the pattern of pronunciation symbols into pronunciation sections for the converted word, each pronunciation section corresponding to a syllable in the converted word;

dividing the pattern of pronunciation symbols into pronunciation sections for each pattern in the database used in the compare; and

comparing the pronunciation sections for the converted word to the corresponding pronunciation sections for each pattern used in the compare; and

providing the word in the list corresponding to the pattern having the match value that is indicative of a match to the converted word.