Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020117605 - EXPEDITING INTERACTION WITH A DIGITAL ASSISTANT BY PREDICTING USER RESPONSES

Note: Text based on automatic Optical Character Recognition processes. Please use the PDF version for legal matters

[ EN ]

CLAIMS

1. One or more computing devices for controlling interaction with a digital assistant, comprising:

hardware logic circuitry, the hardware logic circuitry corresponding to: (a) one or more hardware processors that perform operations by executing machine-readable instructions stored in a memory, and/or (b) one or more hardware logic components that perform operations using a task-specific collection of logic gates, the operations including: receiving a system prompt generated by a digital assistant, the digital assistant generating the system prompt in response to an input command provided by a user via an input device, both the system prompt and the input command being expressed in a natural language;

generating a predicted response based on linguistic content of the system prompt, together with contextual features pertaining to a circumstance in which the system prompt was issued, the predicted response corresponding to a prediction of how the user will respond to the system prompt; and

selecting one or more dialogue actions from a plurality of dialogue actions, based on a confidence value associated with the predicted response, the confidence value describing a degree of confidence of the predicted response,

a first dialogue action involving feeding the predicted response into the digital assistant as a proxy user response to the system prompt, without outputting the predicted response or the system prompt to the user via an output device,

the first dialogue action having an effect of expediting interaction between the user and the digital assistant by not requiring the user to respond to the system prompt,

said generating of the predicted response being based on a data store that contains a plurality of record entries, each record entry containing: a previously-generated system prompt; a set of contextual features pertaining to a circumstance in which the previously-generated system prompt was issued; and a previously-encountered user response to the previously-generated system prompt.

2. The one or more computing devices of claim 1, wherein said selecting chooses the first dialogue action when the confidence value is above a prescribed threshold value.

3. The one or more computing devices of claim 1, wherein the predicted response that is fed back into the digital assistant contains a correction to an inaccurate assumption in the system prompt.

4. The one or more computing devices of claim 1, wherein a second dialogue action

involves:

instructing the digital assistant to send the system prompt to the user via the output device,

the digital assistant subsequently receiving an actual user response to the system prompt.

5. The one or more computing devices of claim 4, wherein the operations further include adding a new record entry to the data store which includes the system prompt, contextual features pertaining to a circumstance in which the system prompt was issued, and the actual user response.

6. The one or more computing devices of claim 1, wherein a third dialogue action involves:

sending a confirmation prompt to the user via the output device, the confirmation prompt asking the user to confirm the predicted response,

the digital assistant subsequently receiving an actual user response to the confirmation prompt.

7. The one or more computing devices of claim 6, wherein the operations further include, upon the user confirming the predicted response:

modifying the actual user response so that it expresses the predicted response, to produce a modified user response, when the actual user response indicates acceptance of the predicted response; and

feeding the modified user response into the digital assistant as a proxy user response.

8. The one or more computing devices of claim 1, wherein said generating of the predicted response operates by:

using the system prompt and the contextual features to find a matching record entry in the data store; and

choosing a previously-encountered user response that is associated with the matching record entry as the predicted response.

9. A method, implemented by one or more computing devices, for controlling interaction with a digital assistant, comprising, in a prediction system:

receiving a system prompt generated by a digital assistant, the digital assistant generating the system prompt in response to an input command provided by a user via an input device, both the system prompt and the input command being expressed in a natural language;

generating a predicted response based on linguistic content of the system prompt, together with contextual features pertaining to a circumstance in which the system prompt was issued, the predicted response corresponding to a prediction of how the user will respond to the system prompt; and

according to a first dialogue action, feeding the predicted response into the digital assistant as a proxy user response to the system prompt,

the prediction system generating the predicted response based on a data store that contains a plurality of record entries, each record entry containing: a previously-generated system prompt; a set of contextual features pertaining to a circumstance in which the previously-generated system prompt was issued; and a previously-encountered user response to the previously-generated system prompt,

the prediction system providing a user personalization function external to one or more skill components provided by the digital assistant.

10. A computer-readable storage medium for storing computer-readable instructions, the computer-readable instructions, when executed by one or more hardware processors, performing a method that comprises:

receiving a system prompt generated by a digital assistant, the digital assistant generating the system prompt in response to an input command provided by a user via an input device, both the system prompt and the input command being expressed in a natural language,

the digital assistant including:

a natural language understanding (NLU) component for interpreting the user input command, to provide an interpreted user command;

a dialogue manager for coordinating with one or more skill components to provide an answer to the interpreted user command; and a natural language generator (NGU) component for generating the system prompt, in the natural language, in reply to the answer; generating a predicted response based on linguistic content of the system prompt, together with contextual features pertaining to a circumstance in which the system prompt was issued, the predicted response corresponding to a prediction of how the user will respond to the system prompt; and

selecting one or more dialogue actions from a plurality of dialogue actions, based on a confidence value associated with the predicted response, the confidence value describing a degree of confidence of the predicted response,

a first dialogue action involving feeding the predicted response into the digital assistant as a proxy user response to the system prompt, without outputting the predicted response or the system prompt to the user via an output device;

a second dialogue action involving instructing the digital assistant to send the system prompt to the user via the output device; and a third dialogue action involving sending a confirmation prompt to the user via the output device, the confirmation prompt asking the user to confirm the predicted response;

said generating of the predicted response being based on a data store that contains a plurality of record entries, each record entry containing: a previously-generated system prompt; contextual features pertaining to a circumstance in which the previously-generated system prompt was issued; and a previously-encountered user response to the previously-generated system prompt.

11. The one or more computing devices of claim 4, wherein said selecting chooses the second dialogue action when the confidence value is below a prescribed threshold value.

12. The one or more computing devices of claim 4, wherein the operations further include sending the predicted response to the digital assistant for use by the digital assistant in interpreting the actual user response.

13. The one or more computing devices of claim 6, wherein the confirmation prompt informs the user that the digital assistant is currently taking an action, has already taken the action, or will take the action.

14. The one or more computing devices of claim 6, wherein the confirmation prompt asks the user to confirm the predicted response before the digital assistant takes an action based on the predicted response.

15. The one or more computing devices of claim 1, wherein said generating of the predicted response involves, at selected times devoted to promoting exploration over exploitation within a solution space that defines a set of possible predicted responses that can be chosen, choosing a predicted response among the set of possible predicted responses, at least one non-chosen predicted response in the set having a higher confidence value than the predicted response that is chosen.