Inferring User Sleep Patterns

• US 20160292584A1
• Filed: 03/31/2015
• Published: 10/06/2016
• Est. Priority Date: 03/31/2015
• Status: Abandoned Application

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Cohn, Ido, Weinberg, Shira, Priness, Ido
• Application Number:
  US14/675,390
• Publication Number:
  US 20160292584A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 50/01 Social networking

Abstract

Methods, computer systems, and computer storage media are provided for inferring sleep-related aspects for a user based, in part, on sensor data reflecting user activity detected by one or more sensors. In an embodiment, a user sleep model is trained using a dataset that includes previously-sensed data, descriptive information associated with the previously-sensed data, and/or interpretive data extracted from the previously-sensed data describing circumstances surrounding users when the data was acquired. In an embodiment, services providing time-sensitive recommendations personalized for a user'"'"'s sleeping pattern using the inferred sleep-related aspects.

Claims

1. A computerized system comprising:

one or more sensors configured to provide sensor data reflecting user activity detected by the one or more sensors and;

a sleep model engine configured to train a user sleep model associated with a user based at least in part on previously-sensed interaction data comprised of sensor data provided at a first time;

one or more processors; and

one or more computer storage media storing computer-useable instructions that, when used by the one or more processors, cause the one or more processors to perform operations comprising;

populating, using the sleep model engine, an interaction dataset based at least in part on the previously-sensed interaction data;

training the user sleep model by analyzing the interaction dataset to identify one or more sleep-related features and sleep-related logic that maps sensor data to the one or more sleep-related features and defines logical relationships between the one or more sleep-related features and sleep-related inferences; and

evaluating currently-sensed interaction data comprised of sensor data provided at a second time subsequent to the first time, with the user sleep model to determine a sleep-related inference.

2. The computerized system of claim 1, further comprising updating a sleep-related profile associated with the user using the sleep-related inference, wherein the sleep-related profile is comprised one or more sleep-related aspects of a sleeping pattern.

3. The computerized system of claim 2, further comprising providing a time-sensitive recommendation to the user based on the one or more sleep-related aspects of the sleep-related profile associated with the user.

4. The computerized system of claim 1, wherein the sensor data includes session data reflecting user online activity.

5. The computerized system of claim 1, wherein the sensor data includes user activity occurring over more than one user device.

6. The computerized system of claim 1, wherein the sensor data includes user data associated with communication events.

7. The computerized system of claim 1, wherein the interaction dataset is populated with previously-sensed interaction data from multiple users associated with a crowd-sourced interaction dataset.

8. The computerized system of claim 7, wherein previously-sensed interaction data from the crowd-sourced interaction dataset is filtered prior to populating the interaction dataset using at least one user attribute filter.

9. The computerized system of claim 1, wherein training the user sleep model is further comprised of analyzing the interaction dataset to assign at least one sleep-related weight for the one or more sleep-related features that reflects a relative statistical significance of corresponding sleep-related features in determining the sleep-related inference.

10. The computerized system of claim 1, further comprising identifying the sleep-related inference as an outlier inference using a pre-determined cutoff.

11. The computerized system of claim 10, wherein data associated with the outlier inference is stored in an outlier dataset, and wherein an alternative user sleep model is trained, in part, using the outlier dataset.

12. The computerized system of claim 11, wherein the alternative user sleep model is adapted to generate alternative sleep-related profiles for the user.

13. A computerized method comprising:

receiving, at a server, a request from a service for a sleep-related aspect of a sleeping pattern of a user, the service for providing time-sensitive recommendations to the user;

in response to receiving the request, identifying the sleep-related aspect using a sleep-related profile associated with the user, the sleep-related profile being generated with sleep-related inferences determined, at least in part, using sensor data reflecting user activity detected by one or more sensors and descriptive information associated with the sensor data; and

providing the identified sleep-related aspect and a confidence score associated with the identified sleep-related aspect to the service, the confidence score quantifying a likelihood of the identified sleep-related aspect coinciding with the user'"'"'s actual sleeping pattern.

14. The computerized method of claim 13, wherein the received request comprises a threshold usability criterion that specifies a minimum confidence score that the requested sleep-related aspect must satisfy to be usable by the service.

15. The computerized method of claim 14, further comprising determining the confidence score associated with the identified sleep-related aspect satisfies the threshold usability criterion prior to providing the identified sleep-related aspect to the service.

16. The computerized method of claim 13, wherein the service is an application on a user device associated with the user, and wherein the identified sleep-related aspect provided by the server is used by the application to personalize time-sensitive recommendations according to the user'"'"'s sleeping pattern.

17. The computerized method of claim 13, wherein the received request comprises request tailoring data adapted to elicit sleep-related aspects for the user only determined using sensor data associated with particular descriptive information.

18. The computerized method of claim 17, further comprising verifying the identified sleep-related aspect was only determined using sensor data associated with particular descriptive information prior to providing the identified sleep-related aspect to the service.

20. The one or more computer storage devices of claim 18, wherein the service associated with the one or more computing devices is a server-side service hosted by a third-party server or a cloud-based service.

19. One or more computer storage devices storing computer-useable instructions that, when used by one or more computing devices, cause the one or more computing devices to perform a method for providing time-sensitive recommendations that are personalized to a sleeping pattern of a user, the method comprising:

transmitting a request, by a service associated with the one or more computing devices, for a sleep-related aspect unaddressed event based at least in part on user data from a user device, the unaddressed event being associated with a user;

in response to the request, receiving the sleep-related aspect for the user, the received sleep-related aspect generated with sleep-related inferences determined, at least in part, using sensor data reflecting user activity detected by one or more sensors and interpretive data extracted from the sensor data that describes circumstances surrounding users when the sensor data was acquired; and

providing time-sensitive recommendations to the user that are personalized according to the sleeping pattern of the user.