Assessing personalized risk for a user on a journey

Patent metadata

Field	Value
Publication number	US-10768002-B2
Application number	US-201715794321-A
Country	US
Kind code	B2
Filing date	Oct 26, 2017
Priority date	Oct 26, 2017
Publication date	Sep 8, 2020
Grant date	Sep 8, 2020

Abstract

A system and method for learning and predicting personalized risk associated with a journey for a user are presented. Contextual data may be gathered and analyzed from a plurality of data sources relating to a journey of a user. A risk associated with the journey may be learned for the user according to the contextual data. One or more risk models may be generated according to the learned risks. One or more potential risks associated with a subsequent journey may be predicted using the one or more risk models.

First claim

The invention claimed is:

1. A method for learning and predicting personalized risk associated with a journey for a user by a processor, comprising:
analyzing contextual data gathered from a plurality of data sources relating to a journey of a user;
learning one or more risks associated with the journey according to the contextual data;
generating one or more risk models associated with the one or more risks; wherein generating the one or more risk models includes detecting an anomaly including a deviation between an expected route and an actual route taken by the user, and collecting feedback information from the user regarding the anomaly for determining the one or more risks and a cause of the one or more risks, the feedback information collected by cognitively interacting with the user via a cognitive component such that the user explains the feedback information, including a reasoning for the anomaly, to the cognitive component using natural language dialog; and
predicting one or more potential risks associated with a subsequent journey using the one or more risk models; wherein the one or more potential risks are automatically presented to the user via a computing device associated with the processor.

2. The method of claim 1 , further including monitoring the journey, the user, one or more alternative users associated with the journey, or combination thereof to collect the contextual data, wherein the contextual data includes traffic data, weather data, road conditions, behavior of the user, a health state of the operator, biometric data of the user, sensor data, learned behavioral data of the user in relation to the journey, learned behavioral data of the one or more alternative users, risks shared from the one or more alternative users, or a combination thereof.

3. The method of claim 1 , wherein the detecting of the anomaly includes:
comparing current behavior of the user with previously observed behavior of the user on the journey to detect the anomaly;
comparing current behavior of the user with behavior of one or more alternative users on the journey to detect the anomaly; or
comparing the current behavior of the user with defined, standardized behavior.

4. The method of claim 1 , further including matching the anomaly with one or more causes of the anomaly from a list of previously received information relevant to the journey.

5. The method of claim 1 , further including learning the one or more risks or the one or more risk models from the user, a plurality of collaborative users on the journey, a machine learning operation, or a combination thereof.

6. The method of claim 1 , further including initializing a machine learning mechanism to learn behavior of the user on the journey and the one or more risk models, wherein the plurality of data sources includes at least sensor based devices associated with the user or vehicle, wearable sensors, camera devices, data sources relating to Internet of Things (IoT) computing networks, governmental entities, commercial entities, or combination thereof.

7. The method of claim 1 , further including:
monitoring the contextual data and each event associated with the journey; and
detecting the one or more risks associated with the journey, the one or more potential risks associated with the subsequent journey, or combination thereof according to the contextual data, one or more identified events, one or more risk models, or combination thereof.

8. A system for learning and predicting personalized risk associated with a journey for a user, comprising:
one or more computers with executable instructions that when executed cause the system to:
analyze contextual data gathered from a plurality of data sources relating to a journey of a user;
learn one or more risks associated with the journey according to the contextual data;
generate one or more risk models associated with the one or more risks; wherein generating the one or more risk models includes detecting an anomaly including a deviation between an expected route and an actual route taken by the user, and collecting feedback information from the user regarding the anomaly for determining the one or more risks and a cause of the one or more risks, the feedback information collected by cognitively interacting with the user via a cognitive component such that the user explains the feedback information, including a reasoning for the anomaly, to the cognitive component using natural language dialog; and
predict one or more potential risks associated with a subsequent journey using the one or more risk models; wherein the one or more potential risks are automatically presented to the user via one of the one or more computers associated with the executable instructions.

9. The system of claim 8 , wherein the executable instructions further monitor the journey, the user, one or more alternative users associated with the journey, or combination thereof to collect the contextual data, wherein the contextual data includes traffic data, weather data, road conditions, behavior of the user, a health state of the operator, biometric data of the user, sensor data, learned behavioral data of the user in relation to the journey, learned behavioral data of the one or more alternative users, risks shared from the one or more alternative users, or a combination thereof.

10. The system of claim 8 ,
wherein the detecting of the anomaly includes:
comparing current behavior of the user with previously observed behavior of the user on the journey to detect the anomaly;
comparing current behavior of the user with behavior of one or more alternative users on the journey to detect the anomaly; or
comparing the current behavior of the user with defined, standardized behavior.

11. The system of claim 8 , wherein the executable instructions further match the anomaly with one or more causes of the anomaly from a list of previously received information relevant to the journey.

12. The system of claim 8 , wherein the executable instructions further:
learn the one or more risks or the one or more risk models from the user, a plurality of collaborative users on the journey, a machine learning operation, or a combination thereof;
initialize the machine learning mechanism to learn behavior of the user on the journey and one or more risk models, wherein the plurality of data sources includes at least sensor based devices associated with the user or vehicle, wearable sensors, camera devices, data sources relating to Internet of Things (IoT) computing networks, governmental entities, commercial entities, or combination thereof;
monitor the contextual data and each event associated with the journey; or
detect the one or more risks associated with the journey, the one or more potential risks associated with the subsequent journey, or combination thereof according to the contextual data, one or more identified events, one or more risk models, or combination thereof.

13. A computer program product for, by a processor, learning and predicting personalized risk associated with a journey for a user, the computer program product comprising a non-transitory computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions comprising:
an executable portion that analyzes contextual data gathered from a plurality of data sources relating to a journey of a user;
an executable portion that learns one or more risks associated with the journey according to the contextual data;
an executable portion that generates one or more risk models associated with the one or more risks; wherein generating the one or more risk mo

Assignees

• IBM

Inventors

• Epperlein Jonathan
• LASSOUED Yassine
• Mevissen Martin
• Monteil Julien
• Russo Giovanni
• Shorten Robert

Classifications

• G01C21/3697 Primary

  Output of additional, non-guidance related information, e.g. low fuel level (G01C21/3679 takes precedence) · CPC title

• G08G1/096827

  where the route is computed onboard · CPC title

• G06Q50/14

  Travel agencies · CPC title

• G06Q10/0639

  Performance analysis of employees; Performance analysis of enterprise or organisation operations · CPC title

• G06Q10/0635

  Risk analysis of enterprise or organisation activities · CPC title