Human-assisted pronunciation generation

• US 8,160,881 B2
• Filed: 12/15/2008
• Issued: 04/17/2012
• Est. Priority Date: 12/15/2008
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Klee, Matthew M., Cross, Karen M., Murphy, Brian J.
• Primary Examiner(s): Mcfadden, Susan
• Application Number:
  US12/334,590
• Publication Number:
  US 20100153115A1
• Time in Patent Office: 1,219 Days
• Field of Search: 704/260
• US Class Current: 704/260
• CPC Class Codes: G10L 13/08 Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination

Abstract

Pronunciation generation may be provided. First, a pronunciation interface may be provided. The pronunciation interface may be configured to display a word and a plurality of alternatives corresponding to a one of a plurality of parts of the word. The plurality of parts may comprise phonemes or syllables of the word. Next, pronunciation data may be received through the pronunciation interface. The pronunciation data may indicate one of the plurality of alternatives. Then a pronunciation of the word may be generated based upon the received pronunciation data. The pronunciation may correspond to the indicated one of the plurality of alternatives. In addition, the pronunciation data may indicate which one of the plurality of parts of the word is stressed. This stress indication may be received in response to a user sliding a user selectable element to indicate which one of the plurality of parts of the word is stressed.

Claims

1. A method for providing pronunciation generation, the method comprising:

providing a pronunciation interface configured to display,

a word, and

a plurality of alternatives corresponding to the word;

receiving pronunciation data through the pronunciation interface, the pronunciation interface being configured to enable a user to manipulate a pronunciation of the word with a manipulation tool; and

generating, by a computing device, a pronunciation of the word based upon the received pronunciation data.

2. The method of claim 1, wherein providing the pronunciation interface configured to display the word comprises providing the pronunciation interface configured to display the word being within a displayed text string.

3. The method of claim 1, wherein providing the pronunciation interface configured to display the plurality of alternatives comprises providing the pronunciation interface configured to display the plurality of alternatives, each of the plurality of alternatives indicating a pronunciation of the word comprising a heterograph.

4. The method of claim 1, wherein providing the pronunciation interface configured to display the plurality of alternatives comprises providing the pronunciation interface configured to display the plurality of alternatives, each of the plurality of alternatives indicating different pronunciations of the word.

5. The method of claim 1, wherein generating the pronunciation comprises generating the pronunciation by one of the following:

an interactive voice response (IVR) system and an automated teller machine (ATM).

6. A computer-readable non-transitory storage medium that stores a set of instructions which when executed perform a method for providing pronunciation generation, the method executed by the set of instructions comprising:

providing a pronunciation interface configured to display,

a word, and

a plurality of alternatives corresponding to a one of a plurality of parts of the word;

receiving pronunciation data through the pronunciation interface, the pronunciation data indicating a one of the plurality of alternatives, the pronunciation interface being configured to manipulate a pronunciation of the word via a manipulation tool; and

generating a pronunciation of the word based upon the received pronunciation data, the pronunciation corresponding to the indicated one of the plurality of alternatives.

7. The computer-readable non-transitory storage medium of claim 6, wherein providing the pronunciation interface configured to display the word comprises providing the pronunciation interface configured to display the word being within a displayed text string.

8. The computer-readable non-transitory storage medium of claim 6, wherein providing the pronunciation interface configured to display the plurality of alternatives corresponding to the one of the plurality of parts of the word comprises providing the pronunciation interface configured to display the plurality of alternatives corresponding to the one of the plurality of parts wherein the plurality of parts comprise syllables of the word.

9. The computer-readable non-transitory storage medium of claim 6, wherein providing the pronunciation interface configured to display the plurality of alternatives corresponding to the one of the plurality of parts of the word comprises providing the pronunciation interface configured to display the plurality of alternatives corresponding to the one of the plurality of parts wherein the plurality of parts comprise phonemes comprising the word.

10. The computer-readable non-transitory storage medium of claim 6, wherein providing the pronunciation interface comprises providing the pronunciation interface configured to display a user selectable element configured to indicate which one of the plurality of parts of the word is stressed.

11. The computer-readable non-transitory storage medium of claim 6, wherein receiving the pronunciation data through the pronunciation interface comprises receiving the pronunciation data indicating which one of the plurality of parts of the word is stressed.

12. The computer-readable non-transitory storage medium of claim 6, wherein receiving the pronunciation data through the pronunciation interface comprises receiving the pronunciation data indicating which one of the plurality of parts of the word is stressed in response to a user sliding a user selectable element to indicate which one of the plurality of parts of the word is stressed.

13. The computer-readable non-transitory storage medium of claim 6, wherein providing the pronunciation interface comprises displaying, on a first manipulation menu, the word and the plurality of parts of the word.

14. The computer-readable non-transitory storage medium of claim 13, wherein providing the pronunciation interface comprises displaying, on a second manipulation menu, the word and the plurality of alternatives corresponding to the one of the plurality of parts of the word in response to a user selecting the one of a plurality of parts of the word from the first manipulation menu.

15. The computer-readable non-transitory storage medium of claim 6, wherein generating the pronunciation of the word comprises generating the pronunciation of the word with an up prosody based upon a context of the generated pronunciation.

16. The computer-readable non-transitory storage medium of claim 6, wherein generating the pronunciation of the word comprises generating the pronunciation of the word with a down prosody based upon a context of the generated pronunciation.

17. The computer-readable non-transitory storage medium of claim 6, wherein generating the pronunciation comprises generating the pronunciation by one of the following:

an interactive voice response (IVR) system and an automated teller machine (ATM).

18. A system for providing pronunciation generation, the system comprising:

a memory storage; and

a hardware processing unit coupled to the memory storage, wherein the hardware processing unit is operative to;

provide a pronunciation interface configured to prompt a user for text data and sound data corresponding to the text data, the pronunciation interface comprising a manipulation tool for altering a pronunciation of at least a portion of the text data;

receive the text data and the sound data through the pronunciation interface;

correlate the text data with the sound data to produce pronunciation data, the pronunciation data indicating how parts of the text data are to be pronounced as indicated by corresponding parts of the sound data; and

generate a pronunciation of the at least a portion of the text data based upon the pronunciation data.

19. The system of claim 18, wherein the hardware processing unit being operative to receive the text data and the sound data through the pronunciation interface comprises the hardware processing unit being operative to:

receive the text data through a word box on an input menu corresponding to the pronunciation interface; and

receive the sound data in response to a user initiating a record button on the input menu.

20. The system of claim 18, wherein the hardware processing unit being operative to generate the pronunciation of the word comprises the hardware processing unit being operative to generate the pronunciation of the word with one of the following:

an up prosody based upon a context of the generated pronunciation and a down porosity based upon the context of the generated pronunciation.