Generating self-support metrics based on paralinguistic information

Patent metadata

Field	Value
Publication number	US-11354754-B2
Application number	US-202016789126-A
Country	US
Kind code	B2
Filing date	Feb 12, 2020
Priority date	Oct 26, 2016
Publication date	Jun 7, 2022
Grant date	Jun 7, 2022

Abstract

Certain aspects of the present disclosure provide techniques for selecting a response to a self-support query. One example method generally includes receiving an audio stream query including spoken content from a user recorded by a mobile device and determining a set of paralinguistic features from the spoken content. The method further includes estimating an emotional state of the user based on the set of paralinguistic features and identifying subject matter of the spoken content in the audio stream query. The method further includes determining two or more query responses corresponding to the subject matter to present to the user and transmitting at least one query response to the mobile device.

First claim

What is claimed is:

1. A method for selecting a response to a self-support query, the method comprising:
receiving a query including content from a user recorded by a mobile device;
estimating a user attribute of the user based on a set of paralinguistic features extracted from the query;
identifying subject matter of the content in the query;
determining two query responses corresponding to the subject matter;
determining, based on the subject matter, one query response from the two query responses to present to the user based on an A/B test metric that includes a favorability score, an uplift metric generated by an uplift model, and the estimated user attribute, wherein the A/B test metric and the uplift metric include instructions to display the one query response;
transmitting the one query response to display on the mobile device.

2. The method of claim 1 , wherein the method further comprises determining the A/B test metric is applicable to the identified subject matter.

3. The method of claim 2 , wherein the method further comprises:
presenting the one query response to the user based on instructions included in the A/B test metric;
evaluating the user's reaction to the one query response; and
assigning a favorability score to the user's reaction.

4. The method of claim 3 , wherein the method further comprises:
including the user's reaction in a set of A/B test results; and
including the estimated user attribute in a corresponding set of estimated attributes.

5. The method of claim 4 , further comprising an uplift model that is trained with the set of A/B test results and the set of estimated attributes, to provide the uplift model metric.

6. The method of claim 5 , wherein the uplift model generates instructions to provide the one query response related to the identified subject matter.

7. The method of claim 1 , wherein the method further comprises:
determining the A/B test metric is not applicable to the identified subject matter; and
determining the one query response based on the uplift model metric and the estimated user attribute of the user.

8. A computing device, comprising:
a processor; and
a memory including computer readable instructions, which, when executed by the processor, cause the computing device to perform a method for selecting a response to a self-support query, the method comprising:
receiving a query including content from a user recorded by a mobile device;
estimating a user attribute of the user based on a set of paralinguistic features extracted from the query;
identifying subject matter of the content in the query;
determining two query responses corresponding to the subject matter;
determining one query response from the two query responses to present to the user based on an A/B test metric that includes a favorability score, an uplift metric generated by an uplift model, and the estimated user attribute, wherein the A/B test metric and the uplift metric include instructions to display the one query response;
transmitting the one query response to display on the mobile device.

9. The computing device of claim 8 , wherein the method further comprises determining the A/B test metric is applicable to the identified subject matter.

10. The computing device of claim 9 , wherein the method further comprises:
presenting the one query response to the user based on instructions included in the A/B test metric;
evaluating the user's reaction to the one query response; and
assigning a favorability score to the user's reaction.

11. The computing device of claim 10 , wherein the method further comprises:
including the user's reaction in a set of A/B test results; and
including the estimated user attribute in a corresponding set of estimated attributes.

12. The computing device of claim 11 , further comprising an uplift model that is trained with the set of A/B test results and the set of estimated attributes, to provide the uplift model metric.

13. The computing device of claim 12 , wherein the uplift model generates instructions to provide the one query response related to the identified subject matter.

14. The computing device of claim 8 , wherein the method further comprises:
determining the A/B test metric is not applicable to the identified subject matter; and
determining the one query response based on the uplift model metric and the estimated user attribute of the user.

15. A non-transitory computer readable medium comprising instructions, which when executed by a processor of a computing device, cause the computing device to perform a method for selecting a response to a self-support query, the method comprising:
receiving a query including content from a user recorded by a mobile device;
estimating a user attribute of the user based on a set of paralinguistic features extracted from the query;
identifying subject matter of the content in the query;
determining two query responses corresponding to the subject matter;
determining one query response from the two query responses to present to the user based on an A/B test metric that includes a favorability score, an uplift metric generated by an uplift model, and the estimated user attribute, wherein the A/B test metric and the uplift metric include instructions to display the one query response;
transmitting the one query response to display on the mobile device.

16. The non-transitory computer readable medium of claim 15 , wherein the method further comprises determining the A/B test metric is applicable to the identified subject matter.

17. The non-transitory computer readable medium of claim 16 , wherein the method further comprises:
presenting the one query response to the user based on instructions included in the A/B test metric;
evaluating the user's reaction to the one query response; and
assigning a favorability score to the user's reaction.

18. The non-transitory computer readable medium of claim 17 , wherein the method further comprises:
including the user's reaction in a set of A/B test results; and
including the estimated user attribute in a corresponding set of estimated attributes.

19. The non-transitory computer readable medium of claim 18 , further comprising an uplift model that is trained with the set of A/B test results and the set of estimated attributes, to provide the uplift model metric.

20. The non-transitory computer readable medium of claim 15 , wherein the method further comprises:
determining the A/B test metric is not applicable to the identified subject matter; and
determining the one query response based on the uplift model metric and the estimated user attribute of the user.

Assignees

• Intuit Inc

Inventors

• Indyk Benjamin
• Podgorny Igor A
• Chan Raymond

Classifications

• G10L25/63

  for estimating an emotional state · CPC title

• G06F16/313

  Selection or weighting of terms for indexing · CPC title

• G10L25/87

  Detection of discrete points within a voice signal · CPC title

• G06Q40/12 Primary

  Accounting · CPC title

• G10L15/22

  Procedures used during a speech recognition process, e.g. man-machine dialogue · CPC title