Method for accessing data via voice

• US 7,493,259 B2
• Filed: 01/04/2002
• Issued: 02/17/2009
• Est. Priority Date: 01/04/2002
• Status: Active Grant

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Siebel Systems Inc. (Oracle Corporation)
• Inventors: Gorman, Richard, Ambrose, Jesse, Jones, Shannon, Haven, Chris, Harb, Joseph
• Primary Examiner(s): Vo; Huyen X.
• Application Number:
  US10/039,118
• Publication Number:
  US 20070198267A1
• Time in Patent Office: 2,601 Days
• Field of Search: 704/257, 704/246, 704/270, 704/270.1, 704/273, 704/275, 704/7, 704/9, 704/231, 704/251, 379/88.22, 379/88.16
• US Class Current: 704/257
• CPC Class Codes: G06F 16/332 Query formulation
• CPC Class Codes: H04M 3/4936 Speech interaction details speech recognition per se G10L15/00

Abstract

A method for providing access to data via a voice interface. In one embodiment, the system includes a voice recognition unit and a speech processing server that work together to enable users to interact with the system using voice commands guided by navigation context sensitive voice prompts, and provide user-requested data in a verbalized format back to the users. Digitized voice waveform data are processed to determine the voice commands of the user. The system also uses a “grammar” that enables users to retrieve data using intuitive natural language speech queries. In response to such a query, a corresponding data query is generated by the system to retrieve one or more data sets corresponding to the query. The user is then enabled to browse the data that are returned through voice command navigation, wherein the system “reads” the data back to the user using text-to-speech (TTS) conversion and system prompts.

Claims

1. A method for accessing data from an enterprise data system via voice input, comprising:

authenticating a login, wherein the authenticating comprises;

querying a database with a voice identifier,

in response to the querying, verifying the voice identifier and receiving a password for the enterprise data system from the database, and

establishing a connection with the enterprise data system using the password for the enterprise data system;

enabling access to a domain of the enterprise data system, wherein each of a plurality of domains of the enterprise data system corresponds to a respective object or type of data;

determining a navigation context;

receiving a navigation command;

updating the navigation context in response to the navigation command;

providing a system prompt based on the navigation context;

receiving a spoken language query corresponding to the navigation context;

converting the spoken language query into a data query and executing the data query to retrieve data that corresponds to the data query in the accessed domain; and

providing feedback data in a verbal format, wherein the feedback data corresponds to data retrieved from the accessed domain and is based, at least in part, on the navigation context, and the providing the feedback data comprises;

generating audio data by performing a text-to-speech conversion on retrieved data; and

generating a verbalized system response by interspersing the audio data with waveform data of prompts.

2. The method of claim 1, wherein the data query includes reference to a unique enterprise data system user identifier such that the data query returns user-specific data.

3. The method of claim 1, wherein the login comprises a unique user identifier and a personal identification number (PIN).

4. The method of claim 1, further comprising:

converting the spoken language query into a data request in an application-readable form;

identifying one or more objects and data criteria corresponding to the spoken language query by processing the data request; and

formulating the data query based, at least in part, on the identified objects and data criteria.

5. The method of claim 4, wherein the enterprise data system includes an object manager and data manager that are used to enable access to data stored in an enterprise database, further comprising:

passing information corresponding to the identified objects and data criteria to the object manager;

formulating a database query based, at least in part, on the objects and data criteria passed to the object manager in consideration of enterprise database schema information available to the data manager;

submitting the database query to the enterprise database;

receiving a result set back from the enterprise database in response to the database query; and

processing the result set to produce the feedback data.

6. The method of claim 5, further comprising:

extracting object data from the result set; and

defining a slotted data string corresponding to a grammatical form in which data are to be presented;

embedding the object data into slots in the slotted data string to produce the feedback data.

7. The method of claim 1, wherein converting the spoken language query into the data query comprises:

receiving voice input as digital waveform data;

passing the digital waveform data to a voice recognition component;

receiving application-readable data from the voice recognition component corresponding to the spoken language query; and

processing the application-readable data to identify data to be retrieved.

8. The method of claim 1, further comprising:

defining a grammar syntax language comprising a plurality of grammars specifying grammatical formatting of legal inputs; and

identifying data to be retrieved by processing voice input in consideration of the grammar syntax language.

9. The method of claim 1, wherein the providing the feedback data comprises:

defining a slotted data string corresponding to a grammatical form in which data are to be presented;

embedding data retrieved from the enterprise data system in slots defined in the slotted data string to form an embedded data text string;

passing the embedded data text string to a text-to-speech conversion component;

receiving digital waveform data from the text-to-speech conversion component corresponding to the embedded data text string;

streaming the digital waveform data to a device that produces an audible sound in response to processing the digital waveform data to produce a verbalized feedback.

10. The method of claim 9, wherein a plurality of slotted data strings are defined, each corresponding to a respective system response, further comprising:

selecting an appropriate slotted data string from among the plurality of slotted data strings based, at least in part, on the navigation context.

11. The method of claim 9, wherein a plurality of slotted data strings are defined, each corresponding to a respective system response, further comprising:

identifying attributes corresponding to data retrieved from the enterprise data system; and

selecting an appropriate slotted data string from among the plurality of slotted data strings based, at least in part, on the identified attributes.

12. The method of claim 1, wherein providing the feedback data corresponding to data retrieved from the enterprise data system in a verbal format comprises:

storing a plurality of prompt audio files, each comprising prompt digital waveform data that when processed produces a verbalized prompt comprising one or more words;

defining a slotted data string specifying a grammatical form in which data are to be presented by identifying prompt audio files to be streamed and specifying where data are to be inserted relative to the identified prompt audio files;

passing data retrieved from the enterprise data system to a text-to-speech conversion component;

receiving text-to-speech (TTS) digital waveform data from the text-to-speech conversion component in response to the passing;

streaming prompt digital waveform data retrieved from the identified prompt audio files and the TTS digital waveform data to a device that produces an audible sound in response to the retrieved prompt digital waveform data and the TTS digital waveform data to produce a verbalized feedback, wherein the streaming is performed according to an order defined by the slotted data string.

13. The method of claim 12, wherein a plurality of slotted data strings are defined, each corresponding to a respective system response, further comprising:

selecting an appropriate slotted data string from among the plurality of slotted data strings based, at least in part, on the navigation context.

14. The method of claim 12, wherein a plurality of slotted data strings are defined, each corresponding to a respective system response, further comprising:

identifying attributes corresponding to data retrieved from the enterprise data system; and

selecting an appropriate slotted data string from among the plurality of slotted data strings based, at least in part, on the identified attributes.

15. The method of claim 1, further comprising:

enabling navigation in the accessed domain through spoken navigation commands.

16. The method of claim 1, wherein:

the plurality of domains of the enterprise data system comprises an opportunities domain, a contacts domain, an accounts domain, a calendar domain, and an employees domain; and

the voice identifier comprises;

a numeric userID, and

a numeric personal identification number.

17. A method for accessing an enterprise data system via a voice communications device, comprising:

enabling a communications connection to a voice access system;

authenticating a login through the communications connection using a user identifier, wherein the authenticating comprises;

querying a database with the user identifier, and

in response to the querying, verifying the user identifier and receiving from the database an enterprise data system log-in data comprising a password for the enterprise data system;

automatically logging into the enterprise data system using the enterprise data system log-in data;

enabling access to a domain of the enterprise system after the logging into the enterprise data system, each of a plurality of domains corresponding to a respective object or type of data;

determining a navigation context;

receiving a navigation command;

updating the navigation context in response to the navigation command;

providing a system prompt based on the navigation context;

enabling a request that a query be performed using a spoken language query corresponding to the navigation context;

converting the spoken language query into a data query and executing the data query to retrieve data that corresponds to the data query in the accessed domain;

providing feedback data in a verbal format via the communications connection, wherein the feedback data corresponds to data retrieved from the accessed domain and is based, at least in part, on the navigation context, and the providing the feedback data comprises;

generating audio data by performing a text-to-speech conversion on retrieved data; and

generating a verbalized system response by interspersing the audio data with waveform data of prompts.

18. The method of claim 17, wherein the voice communications device comprises a telephone, and the authenticating comprises:

verifying a user identifier and a personal identification number (PIN) received from the telephone through the communications connection.

19. The method of claim 18, wherein the verifying comprises:

if the user identifier was input in a verbal form, verifying a verbal form user identifier;

if the user identifier was input via a keypad on the telephone, verifying a tone form user identifier;

if the PIN was input in a verbal form, verifying a verbal form PIN; and

if the PIN was input keypad on the telephone, verifying a tone form PIN.

20. The method of claim 17, further comprising:

converting the spoken language query into a data request in an application-readable form;

identifying one or more objects and data selection criteria corresponding to the spoken language query by processing the data request; and

formulating the data query based, at least in part, on identified objects and data selection criteria.

21. The method of claim 20, wherein the enterprise data system includes an object manager and data manager that are used to enable access to data stored in an enterprise database, further comprising:

passing information corresponding to identified objects and data selection criteria to the object manager;

formulating a database query based, at least in part, on the objects and data selection criteria passed to the object manager in consideration of enterprise database schema information available to the data manager;

submitting the database query to the enterprise database; and

receiving a result set back from the enterprise database in response to the database query.

22. The method of claim 21, wherein use of the object manager and data manager abstracts objects from how data corresponding to the objects are stored in the enterprise database such that a schema of the enterprise database is changeable without requiring changes to a voice access system component that is external to the enterprise data system.

23. The method of claim 17, further comprising:

retrieving data pertaining to a selected object from the enterprise data system through use of the user identifier upon login to the voice access system; and

providing feedback data corresponding to data that are retrieved in a verbal format via the communications connection.

24. The method of claim 17, further comprising:

enabling navigation in the accessed domain using spoken navigation commands.

25. A method for accessing an enterprise data system via a telephone, comprising:

enabling a telephone connection to a voice access system;

authenticating the telephone connection using a user identifier, wherein the authenticating comprises;

querying a database with the user identifier, and

in response to the querying, verifying the user identifier and receiving from the database an enterprise data system log-in data comprising a password for the enterprise data system;

automatically logging into the enterprise data system using the enterprise data system log-in data;

providing a voice user interface that enables;

access to a plurality of domains, and

navigation and querying of data from an accessed domain using spoken navigation and spoken query commands, wherein each of a plurality of domains comprises data corresponding to a respective type of object in the enterprise data system;

determining a navigation context;

receiving a navigation command;

updating the navigation context in response to the navigation command;

providing a system prompt based on the navigation context;

determining a currently accessed domain among the plurality of domains; and

providing feedback data in a verbal format via the telephone connection in response to spoken navigation and spoken query commands, wherein the feedback data is based, at least in part, on the currently accessed domain, and the providing the feedback data comprises;

generating audio data by performing a text-to-speech conversion on retrieved data; and

generating a verbalized system response by interspersing the audio data with waveform data of prompts.

26. The method of claim 25, wherein the voice user interface includes a set of global voice commands that enables a jump from a current domain to a new domain.

27. The method of claim 25, wherein the voice user interface includes voice commands that are context sensitive to a current navigation context, the method further comprising:

enabling navigation to another navigation context from the current navigation context using navigation voice commands that are based, at least in part, on the current navigation context.

28. The method of claim 25, further comprising:

generating a data query to retrieve data from the enterprise data system in response to one of more of the spoken query commands;

returning a plurality of data sets in response to the data query; and

enabling browsing of the plurality of data sets using verbal input.

29. The method of claim 25, further comprising:

maintaining navigation tracking information for that identifies previous navigation locations; and

selecting system prompts based, at least in part, on the navigation tracking information for a user such that the user is presented with a different system prompt if the user has not previously navigated to a current navigation location than the user is presented with if the user has previously navigated to the current navigation location.

30. The method of claim 25, wherein the spoken navigation and spoken query commands includes a query accessing, which comprises a request to retrieve data corresponding to a domain a user is currently that is provided to the enterprise data system and returns a plurality of data sets comprising header data identifying items pertaining to the current domain, the method further comprising:

enabling the user to browse the header data on an item-by-item basis using verbal navigation commands; and

reading the header data corresponding to each item in response to a user navigation to that item.

31. The method of claim 30, further comprising:

enabling the user to request detail information corresponding to an item that is currently being browsed;

retrieving detail information corresponding to the item currently being browsed from the enterprise database; and

reading the detail information to the user via the telephone connection.

32. A system comprising:

an authentication circuit coupled to a database and configured to authenticate a login by;

querying the database with a voice identifier, and

in response to the querying, verifying the voice identifier and receiving a password for an enterprise data system from the database;

a connection circuit configured to establish a connection with the enterprise data system using the password for the enterprise data system;

an interface configured to access to a domain of the enterprise data system, wherein each of a plurality of domains of the enterprise data system corresponds to a respective object or type of data;

a context circuit for determining a navigation context, receiving a navigation command, updating the navigation context in response to the navigation command, and providing a system prompt based on the navigation context;

a query circuit configured to receive a spoken language query corresponding to the navigation context;

a retrieval circuit configured to convert the spoken language query into a data query and execute the data query to retrieve data that corresponds to the data query in the accessed domain; and

a responder circuit configured to provide data corresponding to the retrieved data in a verbal format by;

generating audio data by performing a text-to-speech conversion on retrieved data; and

generating a verbalized system response by interspersing the audio data with waveform data of prompts.

33. A computer readable medium comprising instructions executable on a processor, wherein the instructions are operable to implement each of:

authenticating a login, wherein the authenticating comprises;

querying a database with a voice identifier;

in response to the querying, verifying the voice identifier and receiving a password for an enterprise data system from the database; and

establishing a connection with the enterprise data system using the password for the enterprise data system;

enabling access to a domain of the enterprise data system, wherein each of a plurality of domains of the enterprise data system corresponds to a respective object or type of data;

determining a navigation context;

receiving a navigation command;

updating the navigation context in response to the navigation command;

providing a system prompt based on the navigation context;

receiving a spoken language query corresponding to the navigation context;

converting the spoken language query into a data query and executing the data query to retrieve data that corresponds to the data query in the accessed domain; and

providing feedback data in a verbal format, wherein the providing the feedback data corresponds to data retrieved from the accessed domain and comprises;

generating audio data by performing a text-to-speech conversion on retrieved data; and

generating a verbalized system response by interspersing the audio data with waveform data of prompts.