State of emotion time series

Patent metadata

Field	Value
Publication number	US-11657235-B2
Application number	US-201916553701-A
Country	US
Kind code	B2
Filing date	Aug 28, 2019
Priority date	Sep 13, 2018
Publication date	May 23, 2023
Grant date	May 23, 2023

Abstract

Methods, computer program products, and systems are presented. The methods include, for instance: inputs of emotion time series data of a user and environmental factor data from one or more data collection device for a user assistance service. A baseline emotion time series is generated and an environmental factor that is likely to have affected changes in state of emotion on a subject is identified by regression analysis. An emotion time series model for the identified environmental factor is produced and prediction of a path to attain a target state of emotion at a certain time in the future is made. Recommendation to achieve the target state of emotion is produced based on the predicted path.

First claim

What is claimed is:

1. A computer implemented method, comprising:
obtaining, by one or more processor, inputs of emotion time series data of a user and environmental factor data from one or more data collection device for a user assistance service;
generating, by the one or more processor, a baseline emotion time series based on the inputs from the obtaining;
ascertaining, by the one or more processor, one or more environmental factor affecting states of emotion as represented in the emotion time series by regression analysis;
building, by the one or more processor, an emotion time series model for one of the one or more environmental factor from the ascertaining;
predicting, by the one or more processor, one or more path to a target state of emotion at a target time in the future on the emotion time series model; and
producing, by the one or more processor, a recommendation specifying one or more activity to perform for the user in order to attain the target state of emotion at the target time.

2. The computer implemented method of claim 1 , further comprising:
obtaining feedback on the one or more activity in the recommendation from the user and updated emotion time series data of the user, subsequent to the user had performed the one or more activity according to the recommendation; and
training the emotion time series model with the feedback and the updated emotion time series data from the obtaining.

3. The computer implemented method of claim 1 , the generating comprising:
calculating an emotion score corresponding to each instance of the obtained emotion time series data according to a preconfigured scale.

4. The computer implemented method of claim 1 , the generating comprising
classifying the obtained emotion time series data into a preconfigured category per each basic emotion, by modeling any linguistic expression from the obtained emotion time series data by topic.

5. The computer implemented method of claim 1 , the building comprising:
modeling, by use of probability density function, an average interval between occurrences of events in an environmental factor from the ascertaining;
modeling, by use of standard distribution, a mean time lapse between changes in states of emotion;
formulating, by use of logit function, a log probability of the one or more environmental factor; and
identifying any drift component in a formula for the log probability from the formulating.

6. The computer implemented method of claim 1 , wherein the one or more activity in the recommendation includes, but is not limited to, activities known to promote a relaxed and positive state of emotion, specifically for the user or for general public having demographic similar to the user, and wherein the one or more activity is stored in an emotion knowledge base.

7. The computer implemented method of claim 1 , wherein the emotion time series data include, but are not limited to, emails, chat texts, blood pressure measurements, heartrate measurements, voice stress levels, and facial images, of or authored by the user.

8. The computer implemented method of claim 1 , wherein the environmental factor data includes an application program which the user is using.

9. The computer implemented method of claim 1 , wherein the environmental factor data includes a location of the user.

10. The computer implemented method of claim 1 , wherein the environmental factor data includes an application program which the user is using, a location that the user is using, scheduled events of the user.

11. The computer implemented method of claim 1 , wherein the one or more activity in the recommendation includes, but is not limited to, activities known to promote a relaxed and positive state of emotion, specifically for the user or for general public having demographic similar to the user, wherein the one or more activity is stored in an emotion knowledge base, and wherein the emotion time series data include, emails, chat texts, blood pressure measurements, heartrate measurements, voice stress levels, and facial images, of or authored by the user.

12. The computer implemented method of claim 1 , further comprising:
obtaining feedback on the one or more activity in the recommendation from the user and updated emotion time series data of the user, subsequent to the user had performed the one or more activity according to the recommendation; and
training the emotion time series model with the feedback and the updated emotion time series data from the obtaining, wherein the generating comprises calculating an emotion score corresponding to each instance of the obtained emotion time series data according to a preconfigured scale.

13. The computer implemented method of claim 1 , further comprising:
obtaining feedback on the one or more activity in the recommendation from the user and updated emotion time series data of the user, subsequent to the user had performed the one or more activity according to the recommendation; and
training the emotion time series model with the feedback and the updated emotion time series data from the obtaining, wherein the generating comprises classifying the obtained emotion time series data into a preconfigured category per each basic emotion, by modeling any linguistic expression from the obtained emotion time series data by topic.

14. The computer implemented method of claim 1 , further comprising:
obtaining feedback on the one or more activity in the recommendation from the user and updated emotion time series data of the user, subsequent to the user had performed the one or more activity according to the recommendation; and
training the emotion time series model with the feedback and the updated emotion time series data from the obtaining, wherein the generating comprises calculating an emotion score corresponding to each instance of the obtained emotion time series data according to a preconfigured scale, and classifying the obtained emotion time series data into a preconfigured category per each basic emotion, by modeling any linguistic expression from the obtained emotion time series data by topic.

15. The computer implemented method of claim 1 , further comprising:
obtaining feedback on the one or more activity in the recommendation from the user and updated emotion time series data of the user, subsequent to the user had performed the one or more activity according to the recommendation; and
training the emotion time series model with the feedback and the updated emotion time series data from the obtaining, wherein the generating comprises calculating an emotion score corresponding to each instance of the obtained emotion time series data according to a preconfigured scale, and classifying the obtained emotion time series data into a preconfigured category per each basic emotion, by modeling any linguistic expression from the obtained emotion time series data by topic, wherein the building comprises modeling, by use of probability density function, an average interval between occurrences of events in an environmental factor from the ascertaining; modeling, by use of standard distribution, a mean time lapse between changes in states of emotion;
formulating, by use of logit function, a log probability of the one or more environmental factor; and identifying any drift component in a formula for the log probability from the formulating, wherein the one or more activity in the recommendation includes activities known to promote a relaxed and positive state of emotion, specifically for the user or for general public having demographic similar to the user, and wherein the one or more activity is stored in

Assignees

• IBM

Inventors

• Anders Kelley
• Fox Jeremy R
• Dunne Jonathan
• Harpur Liam S

Classifications

• G06F18/214

  Generating training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title

• G06F40/30 Primary

  Semantic analysis · CPC title

• G06F18/2431

  Multiple classes · CPC title

• G06V40/174

  Facial expression recognition · CPC title

• G06N5/02

  Knowledge representation; Symbolic representation · CPC title