Spelling words using an arbitrary phonetic alphabet

• US 7,143,037 B1
• Filed: 06/12/2002
• Issued: 11/28/2006
• Est. Priority Date: 06/12/2002
• Status: Expired due to Fees

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Chestnut, Kevin L.
• Primary Examiner(s): Chawan, Vijay
• Assistant Examiner(s): SAINT CYR, LEONARD
• Application Number:
  US10/171,327
• Time in Patent Office: 1,630 Days
• Field of Search: 704/251, 704/254
• US Class Current: 704/251
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

Words are spelled by receiving recognizable words from a user of an interactive voice response system. The first letter of each recognizable word is identified, and a spelling is determined based on the first letters of the recognizable words. Statistics for previous users of the interactive voice response system are determined, where the statistics indicate the number of times each of the recognizable words has been used to indicate a letter. The recognizable word that is most commonly used for each letter is identified. The user is prompted with at least two recognizable words that are most commonly used, where each recognizable word corresponds to a different letter. A selection of one of the recognizable words provided to the user is received.

Claims

1. A method for receiving spelling information from a user of an interactive voice response system, comprising:

receiving one or more recognizable words from a user of an interactive voice response system;

for each recognizable word, identifying a first letter of the recognizable word;

determining a spelling based on the first letters of the recognizable words;

determining statistics for a plurality of previous users of the interactive voice response system, the statistics comprising a number of times each one of the recognizable words has been used to indicate a letter;

identifying the recognizable word that is most commonly used for each letter;

prompting the user with at least two recognizable words that are most commonly used, each recognizable word corresponding to a different letter; and

receiving a selection of one of the recognizable words provided to the user.

2. The method of claim 1, further comprising prompting the user for the spelling.

3. The method of claim 2, further comprising:

identifying a category for one or more of the recognizable words received from the user; and

using recognizable words from the category in a subsequent prompt to the user.

4. The method of claim 1, further comprising:

determining that a spoken word received from the user does not match a recognizable word within a specified degree of statistical certainty; and

in response to determining that the spoken word is not recognizable within the specified degree of statistical certainty, prompting the user for a spelling of the spoken word.

5. The method of claim 1, wherein:

the method further comprises receiving one or more spoken letters from the user; and

the spelling is determined based on the spoken letters and the first letters of the recognizable words.

6. The method of claim 1, further comprising:

determining that the spelling of the particular word corresponds to a particular recognizable word; and

prompting the user with the particular recognizable word.

7. An interactive voice response system, comprising:

an interface operable to receive speech from a user;

a memory operable to store a plurality of recognizable words, the memory is further operable to store statistics for a plurality of previous users, the statistics comprising a number of times each one of the recognizable words has been used to indicate a letter; and

a processor operable to;

determine whether a spoken word received from the user corresponds to one of the recognizable words;

determine a first letter for each spoken word that corresponds to one of the recognizable words;

determine a spelling using the first letters; and

determine a particular recognized word to be used in a prompt based on the statistics.

8. The system of claim 7, wherein the memory is further operable to store the spelling determined from the first letters.

9. The system of claim 7, wherein the processor is further operable to communicate the prompt to the user using the interface.

10. The system of claim 9, wherein the processor is further operable to:

determine a category for the spoken word received from the user; and

prompt the user with recognizable words in the category.

11. The system of claim 7, wherein the processor is further operable to:

determine that the spoken word does not correspond to one of the recognizable words within a specified degree of statistical certainty; and

prompt the user to spell the spoken word.

12. The system of claim 7, wherein the interface receives voice signals directly from the public switched telephone network.

13. An interactive voice response system, comprising:

an interface operable to receive speech from a user;

a memory operable to store a plurality of recognizable words, and store statistics tracking a number of times the user submits a word from each available category; and

a processor operable to enter a first mode and a second mode, and the processor operable to;

identify one of the recognizable words corresponding to a spoken word received from the user;

in the first mode, store the received word in the memory; and

in the second mode, store a first letter of the received word in the memory, and determine a category for the spoken word from a plurality of available categories.

14. The system of claim 13, wherein the processor is further operable to transition from the first mode to the second mode in response to being unable to identify the spoken word within a predetermined degree of statistical certainty.

15. The system of claim 13, wherein:

the interface is further operable to receive a command from the user to transition from the first mode to the second mode; and

the processor is further operable to transition into the second mode in response to receiving the command.

16. The system of claim 13, wherein, in the second mode, the processor is further operable to:

identify individual spoken letters received from the user; and

store the identified letters in the memory.

17. Logic embodied in a computer readable medium operable to perform the steps of:

receiving one or more recognizable words from a user of an interactive voice response system;

for each recognizable word, identifying a first letter of the recognizable word;

determining a spelling based on the first letters of the recognizable words;

determining statistics for a plurality of previous users of the interactive voice response system, the statistics comprising a number of times each one of the recognizable words has been used to indicate a letter;

identifying the recognizable word that is most commonly used for each letter;

providing the user with at least two recognizable words that are most commonly used, each recognizable word corresponding to a different letter; and

receiving a selection of one of the recognizable words provided to the user.

18. The logic of claim 17, further operable to perform the step of prompting the user for the spelling.

19. The logic of claim 18, further operable to perform the steps of:

identifying a category for one or more of the recognizable words received from the user; and

using recognizable words from the category in a subsequent prompt to the user.

20. The logic of claim 17, further operable to perform the steps of:

determining that a spoken word received from the user does not match a recognizable word within a specified degree of statistical certainty; and

in response to determining that the spoken word is not recognizable within the specified degree of statistical certainty, prompting the user for a spelling of the spoken word.

21. The logic of claim 17, wherein:

the logic is further operable to perform the steps of receiving one or more spoken letters from the user; and

the spelling is determined based on the spoken letters and the first letters of the recognizable words.

22. The logic of claim 17, wherein the logic is further operable to perform the steps of:

determining that the spelling of the particular word corresponds to a particular recognizable word; and

prompting the user with the particular recognizable word.