System and method for providing and using universally accessible voice and speech data files

• US 6,400,806 B1
• Filed: 04/05/1999
• Issued: 06/04/2002
• Est. Priority Date: 11/14/1996
• Status: Expired due to Term

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Vois Corporation
• Inventors: Uppaluru, Premkumar V.
• Primary Examiner(s): WEAVER, SCOTT LOUIS
• Application Number:
  US09/286,194
• Time in Patent Office: 1,156 Days
• Field of Search: 379/67.1, 379/88.01, 379/88.02, 379/88.04, 379/88.16, 379/88.17, 379/88.19, 379/88.2, 379/88.21, 379/88.22, 379/88.23, 379/88.24, 379/88.27, 379/93.05, 379/201, 704/270, 704/271, 704/272, 704/275, 370/401, 395/200.3, 395/200.33, 395/200.47, 395/200.48, 395/200.49, 395/200.55, 709/217, 709/218, 709/219
• US Class Current: 379/88.02
• CPC Class Codes: H04L 67/02 based on web technology, e.g. hypertext transfer protocol [HTTP]
• CPC Class Codes: H04M 2201/405 involving speaker-dependent recognition adaptation to speaker for speech recognition G10L15/07
• CPC Class Codes: H04M 3/4938 comprising a voice browser which renders and interprets, e.g. VoiceXML

Abstract

A system and method provides universal access to voice-based documents containing information formatted using MIME and HTML standards using customized extensions for voice information access and navigation. These voice documents are linked using HTML hyper-links that are accessible to subscribers using voice commands, touch-tone inputs and other selection means. These voice documents and components in them are addressable using HTML anchors embedding HTML universal resource locators (URLs) rendering them universally accessible over the Internet. This collection of connected documents forms a voice web. The voice web includes subscriber-specific documents including speech training files for speaker dependent speech recognition, voice print files for authenticating the identity of a user and personal preference and attribute files for customizing other aspects of the system in accordance with a specific subscriber.

Claims

1. A method for delivering caller-customized services to a telephone caller, comprising:

storing caller-specific information in a computer file on a computer network in accordance with a universal resource locator (URL) address wherein the stored caller-specific information includes a master voice signature for the caller;

prompting the caller to input identifying information;

responsive to the identifying information, determining a URL for the file storing the caller-specific information;

retrieving the caller-specific information from the file stored at the URL; and

accessing information in a voice web in accordance with the caller-specific information wherein accessing information in a voice web in accordance with the caller-specific information comprises;

prompting the caller for a voice signature,

recording the voice signature, and

comparing the voice signature to the recorded voice signature to determine whether there is a match.

2. A method for delivering caller-customized services to a telephone caller, comprising:

storing caller-specific information in a computer file on a computer network in accordance with a universal resource locator (URL) address wherein the stored caller-specific information includes a speaker dependent speech recognition training file for the caller;

prompting the caller to input identifying information;

responsive to the identifying information, determining a URL for the file storing the caller-specific information;

retrieving the caller-specific information from the file stored at the URL; and

accessing information in a voice web in accordance with the caller-specific information wherein accessing information in a voice web in accordance with the caller-specific information comprises;

prompting the caller for voice commands,

recording the voice commands, and

performing speaker dependent speech recognition on the voice commands using the training file for the caller.

3. In a computer system coupled to a computer network, wherein the computer network is the Internet, a method of providing user specific input to a computer program, comprising:

determining a universal resource locator (URL) address corresponding to a user;

retrieving, over the computer network, a personal profile associated with the user wherein the personal profile includes data for voice authentication and is stored at the determined URL address;

accessing information included in the personal profile to affect the execution of a computer program for navigating and accessing information in a voice web;

receiving a user authentication request;

retrieving user authentication data from the personal profile;

collecting voice data from the user;

processing the collected voice data; and

comparing the processed voice data to the authentication data to authenticate the identity of the user.

4. The method of claim 3 wherein collecting voice data from the user includes collecting voice data from the user using a telephone.

5. In a computer system coupled to a computer network, wherein the computer network is the Internet, a method of providing user specific input to a computer program, comprising:

determining a universal resource locator (URL) address corresponding to a user;

retrieving, over the computer network, a personal profile associated with the user wherein the personal profile includes data for speaker dependent speech recognition and is stored at the determined URL address;

accessing information included in the personal profile to affect the execution of a computer program for navigating and accessing information in a voice web;

receiving a voice command from the user;

performing speaker dependent speech recognition to identify the voice command; and

executing the recognized voice command.

6. The method of claim 5 wherein receiving a voice command from the user includes receiving a voice command from the user using a telephone.

7. A speech processing system, comprising:

a computer network;

a gateway computer coupled to the computer network adapted to receive subscriber commands;

a server computer program coupled to the network;

a user profile stored on the computer network;

voice web pages stored on the computer network wherein each voice web page is addressable by a universal resource locator (URL) address unique within the computer network and wherein each voice web page includes voice information; and

speech processing software adapted to operate in the computer network for

receiving a user identifier,

receiving a command,

determining a URL address associated with a voice web page responsive to the command,

determining a URL address associated with the user profile responsive to the user identifier,

retrieving the user profile,

retrieving the voice web page, and

generating an output responsive to the user command and information included in the retrieved voice web page and the user profile.

8. The system of claim 7 wherein the computer network is an internet.

9. The system of claim 8 wherein the user profile includes voice print information and wherein the command received by the speech processing software is a command to authenticate the identity of a user.

10. The system of claim 8 wherein the user profile includes speech training information and wherein the command received by the speech processing software is a digitized version of a spoken command and wherein the digitized version is processed using retrieved speech training information.

11. The system of claim 8 further comprising a database query form customized in accordance with at least a portion of the user profile.

12. The system of claim 8 wherein the speech processing software is further adapted to perform the method comprising:

searching a database to return a query result; and

presenting the query result responsive to at least a portion of the user profile.

13. The system of claim 8 further comprising a user directory having a plurality of entries, each entry corresponding to a user identifier and each entry being mapped to a URL address and wherein determining the URL address associated with a voice document responsive to the user identifier and the command includes retrieving a URL address from the user directory.

14. The system of claim 8 wherein the computer network is an internet.

15. The system of claim 14 wherein the user profile includes voice signature information and wherein the command received by the speech processing software is a command to authenticate the identity of a user.

16. The system of claim 14 wherein the user profile includes speech training information and wherein the command received by the speech processing software is a digitized version of a spoken command and wherein the digitized version is processed using retrieved speech training information.

17. The system of claim 14 further comprising a database query form customized in accordance with at least a portion of the user profile.

18. The system of claim 14 wherein the speech processing software is further adapted to perform the method comprising:

searching a database to return a query result; and

presenting the query result responsive to at least a portion of the user profile.

19. The system of claim 14 further comprising a user directory having a plurality of entries, each entry corresponding to a user identifier and each entry being mapped to a URL address and wherein determining the URL address associated with a voice document responsive to the user identifier and the command includes retrieving a URL address from the user directory.

20. A personal voice web for a subscriber comprising:

a plurality of linked voice web pages, each page including an agent for performing various processing tasks required for each respective page and a specially tagged set of key words and touch tone sequences that are associated with embedded anchors and links used for navigation within the web;

each voice web page having access to a respective speech training profiles web page, the speech training profiles web page comprising subscriber specific profiles, the profiles including component sets of related words likely to occur in combination within the respective voice web page, and each voice web page having access to an attributes and preferences web page having access to subscriber specific attributes and preferences specific to the respective voice web page; and

said plurality of linked voice web pages including a personal profile page and service pages.

21. The personal voice web of claim 20 wherein each of said service pages references a service agent communicatively coupled to a service profile agent, wherein the service agent, in response to a subscriber'"'"'s request for its corresponding service, retrieves information from a service database, retrieves from the service profile agent subscriber specific, service specific, speech training profiles and subscriber specific, service specific attributes and preferences, and customizes voice web pages using the retrieved speech training profiles and attributes and preferences for presentation to the subscriber.

22. The personal voice web of claim 21 wherein the service pages include a calendar service page.

23. The personal voice web of claim 21 wherein the service pages include an address book service page.

24. The personal voice web of claim 21 wherein the service pages include an electronic mail service page.

25. The personal voice web of claim 21 wherein the service pages include a stock portfolio service page.

26. The personal voice web of claim 21 wherein the service pages include a news headlines service page.

27. In a personal voice web comprising a plurality of linked voice web pages including a personal profile page and service pages, each voice web page having access to a respective speech training profiles web page, the speech training profiles web page comprising subscriber specific profiles, the profiles including component sets of related words likely to occur in combination within the respective voice web page, and each voice web page having access to an attributes and preferences web page having access to subscriber specific attributes and preferences specific to the respective voice web page, a method for providing customized interaction with a subscriber in response to a request for a service from the subscriber, the method comprising:

responsive to the request for the service, retrieving information from a service database comprising information of the requested service;

retrieving subscriber specific speech training profiles and subscriber specific attributes and preferences applicable to the requested service; and

customizing voice web pages in accordance with the subscriber specific speech training profiles and attributes and preferences applicable to the requested service for presentation to the subscriber.

28. The method of claim 27 wherein the service database includes a calendar service database.

29. The method of claim 27 wherein the service database includes an address book service database.

30. The method of claim 27 wherein the service database includes an electronic mail service database.

31. The method of claim 27 wherein the service database includes a stock portfolio service database.

32. The method of claim 27 wherein the service database includes a news headlines service database.