Voice and telephone keypad based data entry for interacting with voice information services

• US 7,447,299 B1
• Filed: 07/24/2000
• Issued: 11/04/2008
• Est. Priority Date: 05/02/2000
• Status: Expired due to Fees

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Partovi, Hadi, Alpert, David Bryan, Brathwaite, Roderick Steven, Arons, Barry, Belldina, Jeremy
• Primary Examiner(s): Anwah; Olisa
• Application Number:
  US09/621,715
• Time in Patent Office: 3,025 Days
• Field of Search: 379/88.01, 379/93.18, 379/93.01, 379/88.16, 379/88.24
• US Class Current: 379/88.01
• CPC Class Codes: G06F 16/3322 using system suggestions G06F16/3325 takes precedence
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: H04M 2203/251 where a voice mode or a visual mode can be used interchangeably
• CPC Class Codes: H04M 3/493 Interactive information services, e.g. directory enquiries ; Arrangements therefor, e.g. interactive voice response [IVR] systems or voice portals

Abstract

This invention relates in general to methods for receiving and interpreting an alphanumeric input string entered on a telephone keypad using one keystroke per input string character. The set of matching objects in a database are selected and sorted and presented to the user as an ordered list from which the user may select the desired item on the list through a further numeric keystroke entry or through a spoken selection that is matched against a restricted voice recognition grammar for improved recognition accuracy. In an alternative embodiment, the matching objects in the database are selected and sorted and the response associated with each matched object is presented in sorted sequence.

Claims

1. A method in a voice response system of receiving input of a keyword, the method comprising:

providing a list of keywords of characters;

providing a list of words, each word having an output utterance that is an audio representation of the word, each word being associated with a keyword in the provided list of keywords, and each word having a different spelling from its associated keyword;

receiving from a user a key sequence with a key of the key sequence representing multiple characters;

identifying from the received key sequence without other input from the user those keywords of the list of keywords whose initial characters match the possible characters of the received key sequence;

after identifying the keywords of the list that match,

outputting an output utterance corresponding to the words associated with each identified keywords; and

prompting the user to select an identified keyword by speaking the output utterance corresponding to the word associated with the keyword to be selected; and

after outputting the output utterances, inputting from the user an input utterance corresponding to one of the output utterances; and

recognizing the input utterance using a constrained recognition grammar that is constrained by the output utterances corresponding to the words associated with the identified keywords such that the input utterance can only be recognized as a word associated with one of the identified keywords.

2. The method of claim 1 wherein the key sequence is a dual tone multi-frequency key sequence.

3. The method of claim 1 wherein the utterances of the identified keywords are output in an order based on a weighting factor.

4. The method of claim 3 wherein the weighting factor is based on an expected likelihood of the utterance being selected by the user.

5. The method of claim 3 wherein the weighting factor is based on access frequency associated with the utterances.

6. The method of claim 1 wherein the key sequence is a dual tone multi-frequency key sequence, and wherein the utterances of the identified keywords are output in an order based on a weighting factor.

7. The method of claim 1 wherein the inputting from the user a selection of one of the utterances includes the user speaking the selected utterance.

8. The method of claim 1 wherein the inputting from the user a selection of one of the utterances includes the user speaking an alphanumeric character associated with an utterance.

9. The method of claim 1 wherein the inputting from the user a selection of one of the utterances includes receiving from the user a selection of a key corresponding to the utterance.

10. A computer-readable medium encoded with instructions for controlling a voice response system to receive input of a word, by a method comprising:

providing a list of words, each word having an output utterance that is an audio representation of the word;

providing a list of keywords of characters, each keyword associated with a word in the provided list of words and each keyword having a different spelling from its associated word;

receiving from the user a key sequence with a key of the key sequence representing multiple characters, each key represented as a dual tone multi-frequency key;

identifying from the received key sequence without other input from the user those keywords of the list whose initial characters match the possible characters of the received key sequence;

after identifying the keywords of the list that match,

outputting an output utterance corresponding to the word associated with the identified keywords in an order based on a weighting factor for the utterances; and

prompting the user to select an identified keyword by speaking the output utterance corresponding to the word associated with the keyword to be selected; and

after outputting the output utterances, inputting from the user an input utterance corresponding to one of the output utterances; and

recognizing the input utterance using a constrained recognition grammar that is constrained by the output utterances corresponding to the words associated with the identified keywords such that the input utterance can only be recognized as a word associated with one of the identified keywords.

11. The computer-readable medium of claim 10 wherein the weighting factor is based on an expected likelihood of the utterance being selected by the user.

12. The computer-readable medium of claim 10 wherein the weighting factor is based on access frequency associated with the utterances.

13. The computer-readable medium of claim 10 wherein the inputting from the user a selection of one of the utterances includes the user speaking the selected utterance.

14. The computer-readable medium of claim 10 wherein the inputting from the user a selection of one of the utterances includes the user speaking an alphanumeric character associated with an utterance.

15. The computer-readable medium of claim 10 wherein the inputting from the user a selection of one of the utterances includes receiving from the user a selection of a key corresponding to the utterance.

16. A voice response system that receives input of a word from a user, comprising:

a component that provides a list of words and an output utterance for each of the words;

a component that provides a list of keywords of characters, each keyword associated with a word in the provided list of words and each keyword having a different spelling from its associated word;

a component that receives from a user a key sequence with a key of the key sequence representing multiple characters;

a component that identifies from the received key sequence without other input from the user those keywords of the list of keywords whose initial characters match the possible characters of the received key sequence;

a component that, after identifying the keywords of the list that match, outputs an output utterance corresponding to words associated with each of the identified keywords and prompts the user to select a word by speaking the output utterance corresponding to the word to be selected;

a component that, after outputting the output utterances, inputs from the user an input utterance corresponding one of the output utterances; and

a component that recognizes the input utterance using a constrained recognition grammar that is constrained by the output utterances corresponding to the words associated with the identified keywords such that the input utterance can only be recognized as a word associated with one of the identified keywords.

17. The system of claim 16 wherein the utterances are output before the prompting.

18. The system of claim 16 wherein the utterances are output after the prompting.