Natural language voice assistant

• US 10,438,583 B2
• Filed: 07/20/2016
• Issued: 10/08/2019
• Est. Priority Date: 07/20/2016
• Status: Active Grant

• Current Assignee: Lenovo PC International Limited (Lenovo Group Ltd.)
• Original Assignee: Lenovo Singapore Pte Limited (Lenovo Group Ltd.)
• Inventors: Hunt, James Anthony, VanBlon, Russell Speight, Mese, John Carl, Peterson, Nathan J.
• Primary Examiner(s): Kim, Jonathan C
• Application Number:
  US15/215,099
• Publication Number:
  US 20180025724A1
• Time in Patent Office: 1,175 Days
• Field of Search: None
• CPC Class Codes: G06F 16/243 Natural language query formulation
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

A voice assistant of a device is activated not by a key word being spoken but by recognizing speech and determining whether context of the speech indicates that audible voice assistance is appropriate.

Claims

1. A device, comprising:

at least one digital processor; and

storage accessible to the at least one digital processor and comprising instructions executable by the at least one digital processor to;

receive speech from a user;

without receiving a user command to enter voice recognition mode, execute voice recognition on the speech to return plural words;

using the plural words as entering argument, access at least one database to correlate the plural words to ancillary information;

responsive to identifying that at least a first word in the plural words is correlated in the at least one database to first ancillary information and responsive to identifying a first confidence level with the first word, return the first ancillary information to the user at least in part by immediately presenting the first ancillary information without further user input; and

responsive to identifying a second confidence level with a second word in the plural words that correlates to second ancillary information different from the first ancillary information, passively present the second ancillary information such that the second ancillary information is not immediately presented, the first confidence level being higher than the second confidence level, wherein the passive presentation of the second ancillary information comprises collecting different pieces of ancillary information over time based on different respective portions of the received speech and periodically presenting the different pieces of the ancillary information together every N minutes to allow the user to determine whether to use the different pieces of ancillary information.

2. The device of claim 1, comprising at least one audio speaker, and wherein at least some of the ancillary information is output on the at least one audio speaker.

3. The device of claim 1, wherein the instructions are executable by the at least one digital processor to:

using the plural words as entering argument, access a grammar database;

determine, using the plural words, whether the grammar database indicates at least one word is indicated; and

responsive to the grammar database indicating at least one word is indicated, return the ancillary information, the ancillary information being related to the at least one word.

4. The device of claim 1, wherein the instructions are executable by the at least one digital processor to:

using the plural words as entering argument, access the at least one database;

determine, using the plural words, whether the at least one database indicates additional information is associated with the plural words; and

responsive to the at least one database indicating additional information is associated with the plural words, return the ancillary information, the ancillary information comprising at least some of the additional information.

5. The device of claim 1, wherein the at least one database comprises plural phrase entries correlated with respective ancillary information, at least a first phrase entry being correlated to more than one ancillary information.

6. The device of claim 1, wherein return of the ancillary information comprises collecting respective pieces of ancillary information at different times based on different respective portions of the received speech that are received at different times and then presenting together, on a display, the different pieces of ancillary information.

7. The device of claim 6, wherein the different pieces of ancillary information are presented on the display together via a graphical user interface (GUI), the GUI comprising respective indications adjacent to the respective pieces of ancillary information that specify when the different respective portions of speech were received.

8. The device of claim 1, wherein the device customizes the at least one database over time based on one or more phrases from a particular user that are identified as spoken more than once, and wherein the at least one database is customized to include the one or more phrases identified as spoken more than once.

9. The device of claim 8, wherein the at least one database is customized to omit other phrases that have not been identified as spoken by the particular user more than once.

10. The device of claim 8, wherein the one or more phrases are single-purpose phrases.

11. The device of claim 8, wherein the one or more phrases are identified as spoken more than a threshold number of times greater than one.

12. The device of claim 8, wherein the at least one database is customized to omit other phrases in dialects outside of the particular user'"'"'s dialect.

13. A computer readable storage medium (CRSM) that is not a transitory signal, the computer readable storage medium comprising instructions executable by at least one digital processor of a device to:

receive speech from a user;

without receiving a user command to enter voice recognition mode, execute voice recognition on the speech to return plural words;

using the plural words as entering argument, access at least one database to correlate the plural words to ancillary information;

responsive to identifying that at least a first word in the plural words is correlated in the at least one database to first ancillary information and responsive to identifying a first confidence level with the first word, return the first ancillary information to the user at least in part by immediately presenting the first ancillary information without further user input; and

responsive to identifying a second confidence level with a second word in the plural words that correlates to second ancillary information different from the first ancillary information, passively present the second ancillary information such that the second ancillary information is not immediately presented, the first confidence level being higher than the second confidence level, wherein the passive presentation of the second ancillary information comprises collecting different pieces of ancillary information over time based on different respective portions of the received speech and periodically presenting the different pieces of the ancillary information together every N minutes to allow the user to determine whether to use the different pieces of ancillary information.

14. The CRSM of claim 13, wherein the instructions are executable by the at least one digital processor to:

using the plural words as entering argument, access a grammar database;

determine, using the plural words, whether the grammar database indicates at least one word is indicated; and

responsive to the grammar database indicating at least one word is indicated, return the ancillary information, the ancillary information being related to the at least one word.

15. The CRSM of claim 13, wherein the at least one database comprises at least a first phrase entry correlated to more than one ancillary information.

16. A computer-implemented method, comprising:

receiving speech from a user;

executing, without receiving a user command to enter voice recognition mode, voice recognition on the speech to return plural words;

accessing, using the plural words as entering argument, at least one database to correlate the plural words to ancillary information;

responsive to identifying that at least a first word in the plural words is correlated in the at least one database to first ancillary information and responsive to identifying a first confidence level with the first word, returning the first ancillary information to the user at least in part by immediately presenting the first ancillary information without further user input; and

responsive to identifying a second confidence level with a second word in the plural words that correlates to second ancillary information different from the first ancillary information, passively presenting the second ancillary information such that the second ancillary information is not immediately presented, the first confidence level being higher than the second confidence level, wherein the passively presenting the second ancillary information comprises collecting different pieces of ancillary information over time based on different respective portions of the received speech and periodically presenting the different pieces of the ancillary information together every N minutes to allow the user to determine whether to use the different pieces of ancillary information.

17. The computer-implemented method of claim 16, wherein the at least one database comprises at least a first phrase entry correlated to more than one ancillary information.

18. The computer-implemented method of claim 16, comprising:

periodically presenting the different pieces of ancillary information together at least in part by presenting the different pieces of ancillary information together via a user interface (UI) that is presented on a display.

19. The computer-implemented method of claim 16, wherein returning of the ancillary information comprises collecting respective pieces of ancillary information at different times based on different respective portions of the received speech that are received at different times and then presenting, on a display together, the different pieces of ancillary information.

20. The computer-implemented method of claim 19, wherein the different pieces of ancillary information are presented on the display together via a graphical user interface (GUI), the GUI comprising respective indications adjacent to the respective pieces of ancillary information that specify when the different respective portions of speech were received.