Inferring venue visits using semantic information

• US 10,185,973 B2
• Filed: 04/07/2015
• Issued: 01/22/2019
• Est. Priority Date: 04/07/2015
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Priness, Ido, Somech, Haim, Miller, Adi L., Sterenberg, Michael, Vainas, Oded
• Primary Examiner(s): Detweiler, James M
• Application Number:
  US14/680,762
• Publication Number:
  US 20160300263A1
• Time in Patent Office: 1,386 Days
• Field of Search: None
• CPC Class Codes: G06Q 30/0255 based on user history
• CPC Class Codes: G06Q 30/0261 based on user location
• CPC Class Codes: G06Q 30/0267 Wireless devices
• CPC Class Codes: H04W 4/02 Services making use of location information
• CPC Class Codes: H04W 4/029 Location-based management or tracking services

Abstract

A method for inferring venue visits using semantic information includes receiving sensor data from sensors. An indication of a location is received that is associated with a user and determined based on the sensor data. A set of candidate venues associated with the location is determined based on the indication of the location. Sets of semantic information associated with the set of candidate venues are determined based on the sensor data. Candidate venues of the set are ranked by confidence that a given candidate venue corresponds to a visited venue of a venue visit based on the set of semantic information associated with the given candidate venue and additional semantic information associated with the user. A highest ranked candidate venue is selected as the visited venue and an indication is provided to a service causing content to be presented to the user based on the selected visited venue.

Claims

1. A computerized system comprising:

one or more sensors configured to provide sensor data;

one or more processors; and

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

extracting semantic information comprising characteristics of one or more users and characteristics of one or more venues from the sensor data, each characteristic associated with a corresponding one of the users or a corresponding one of the venues;

determining a set of candidate venues based on proximities between respective candidate venues and a location associated with a first user;

inferring that a venue visit by the first user occurred based on

an analysis of spatial-temporal data from the sensor data that corresponds to one or more locations associated with the first user,

the extracted semantic information, and

a particular pattern formed by historical values of a tracked variable;

based on the inferring that the venue visit by the first user occurred, generating respective confidence scores for each candidate venue of the set of candidate venues using a probabilistic model that is fed the tracked variable, wherein the tracked variable includes at least one or more user characteristics of the extracted characteristics of users that are associated with the first user and one or more venue characteristics of the extracted characteristics of venues that are associated with the respective candidate venue, wherein each respective confidence score quantifies a confidence that the respective candidate venue is a visited venue of the venue visit by the first user;

ranking the set of candidate venues by their respective confidence scores;

selecting a highest ranked candidate venue of the set of candidate venues as the visited venue for the venue visit using the ranking;

after the selecting the highest ranked candidate venue, using a state-based probabilistic model to perform an optimization of the ranked set of candidate venues across a sequence of tracked venue visits corresponding to a known routine, each of the tracked venue visits in the sequence having corresponding ranked candidate venues, the optimization comprising reducing the respective confidence score of a particular candidate venue of the set of candidate venues based on a highest ranked candidate venue for a different venue visit in the sequence having a same venue type as the particular candidate venue;

adjusting the ranking of at least some of the set of candidate venues based on the reduced respective confidence score;

selecting a new highest ranked candidate venue of the set of candidate venues as the visited venue for the venue visit using the adjusted ranking; and

providing, to a service, an indication of the selected visited venue that corresponds to the new highest ranked candidate venue, the indication causing content to be presented to the first user based on the selected new highest ranked visited venue.

2. The computerized system of claim 1, wherein the sensor data comprises location samples of a mobile device associated with the first user, and the proximities are determined with respect to the location of the first user that is inferred from the location samples.

3. The computerized system of claim 1, wherein in the generating respective confidence scores, the confidence for a respective candidate venue increases based on matching the one or more user characteristics associated with the first user to the one or more venue characteristics associated with the respective candidate venue.

4. The computerized system of claim 1, wherein the sequence of tracked venue visits includes at least one venue visit prior to the venue visit and at least one venue visit subsequent the venue visit.

5. The computerized system of claim 1, wherein the known routine has been tracked with respect to the first user over multiple instances of the routine.

6. The computerized system of claim 1, wherein a user characteristic of the one or more user characteristics corresponds to the first user using an application, and the confidence for a respective candidate venue is increased based on a venue characteristic of the one or more venue characteristics associating the application with the respective candidate venue.

7. The computerized system of claim 1, wherein the one or more user characteristics are inferred from a search history associated with the first user or web browsing history associated with the first user.

8. The computerized system of claim 1, wherein at least one candidate venue in the set of candidate venues is a categorical venue and at least another candidate venue in the set of candidate venues is a specific venue.

9. The computerized system of claim 1, wherein one of the one or more venue characteristics corresponds to an event associated with a respective candidate venue and the confidence for the respective candidate venue increases based on determining that the venue visit corresponds to the event at the respective candidate venue.

10. The computerized system of claim 1, wherein the inferring that the venue visit by the first user has occurred is based on at least a prior tracked venue visit, and at least one subsequent tracked venue visit.

11. A computerized method comprising:

receiving, by one or more processors, sensor data from one or more sensors;

determining, by the one or more processors, an indication of a location, the location being associated with a user and determined based at least in part on the sensor data;

determining, by the one or more processors and from a particular pattern formed by historical values of a tracked variable, a set of candidate venues associated with the location;

extracting, by the one or more processors and based at least in part on the sensor data, sets of semantic information associated with the set of candidate venues, each set being associated with a respective candidate venue in the set of candidate venues, the sets of semantic information comprising characteristics of one or more users and characteristics of one or more venues;

inferring, by the one or more processors, that a venue visit by the user occurred based on

an analysis of spatial-temporal data from the sensor data that corresponds to one or more locations associated with the user

the extracted sets of semantic information, and

the particular pattern formed by historical values of the tracked variable;

generating, by the one or more processors, respective confidence scores for each candidate venue of the set of candidate venues using a probabilistic model that is fed the tracked variable, wherein the tracked variable includes at least one or more user characteristics of the extracted characteristics of users that are associated with the user and one or more venue characteristics of the extracted characteristics of venues that are associated with the respective candidate venue, wherein each respective confidence score quantifies a confidence that the respective candidate venue is a visited venue of the venue visit by the user;

ranking, by the one or more processors, at least some of the set of candidate venues by their respective confidence score;

using a state-based probabilistic model, by the one or more processors, to perform an optimization of the ranked set of candidate venues across a sequence of tracked venue visits corresponding to a known routine, each of the tracked venue visits in the sequence having corresponding ranked candidate venues, the optimization comprising reducing the respective confidence score of a particular candidate venue of the set of candidate venues based on a highest ranked candidate venue for a different venue visit in the sequence having a same venue type as the particular candidate venue;

adjusting, by the one or more processors, the ranking of at least some of the set of candidate venues based on the reduced respective confidence score;

selecting, by the one or more processors, a new highest ranked candidate venue as the visited venue of the venue visit using the adjusted ranking; and

providing, by the one or more processors and to a service, an indication of the selected visited venue that corresponds to the new highest ranked candidate venue, the indication causing content to be presented to the user based on the selected new highest ranked visited venue.

12. The computerized method of claim 11, wherein the sensor data comprises location samples of a mobile device associated with the user, and the indication of the location is determined based at least in part on the location samples.

13. The computerized method of claim 11, further comprising determining, by the one or more processors, venue pattern information from the sensor data, at least some of the semantic information in the sets of semantic information being inferred from the venue pattern information.

14. The computerized method of claim 11, further comprising determining, by the one or more processors, user pattern information from the sensor data, at least some of the semantic information being inferred from the user pattern information.

15. The computerized method of claim 11, wherein the confidence for the given candidate venue is increased based on matching one or more user characteristics in the semantic information associated with the user to one or more venue characteristics in the semantic information associated with the given candidate venue.

16. A computerized method comprising:

receiving, at a server, first one or more network communications comprising sensor data from one or more sensors;

determining, by the server, a location associated with a user based at least in part on the sensor data;

determining, by the server and from a particular pattern formed by historical values of a tracked variable, a set of candidate venues associated with the location;

extracting, by the server and based at least in part on the sensor data, sets of semantic information associated with the set of candidate venues, each set being associated with a respective candidate venue in the set of candidate venues, the sets of semantic information comprising characteristics of one or more users and characteristics of one or more venues;

inferring, by the server, that a venue visit by the user occurred based on

an analysis of spatial-temporal data from the sensor data that corresponds to one or more locations associated with the user

the extracted sets of semantic information, and

the particular pattern formed by historical values of the tracked variable;

generating, by the server, respective confidence scores for each candidate venue of the set of candidate venues using a probabilistic model that is fed the tracked variable, wherein the tracked variable includes at least one or more user characteristics of the extracted characteristics of users that are associated with the user and one or more venue characteristics of the extracted characteristics of venues that are associated with the respective candidate venue, wherein each respective confidence score quantifies a confidence that the respective candidate venue is a visited venue of the venue visit by the user;

ranking, by the server, at least some of the set of candidate venues by their respective confidence score;

using a state-based probabilistic model, by the server, to perform an optimization of the ranked set of candidate venues across a sequence of tracked venue visits corresponding to a known routine, each of the tracked venue visits in the sequence having corresponding ranked candidate venues, the optimization comprising reducing the respective confidence score of a particular candidate venue of the set of candidate venues based on a highest ranked candidate venue for a different venue visit in the sequence having a same venue type as the particular candidate venue;

adjusting, by the server the ranking of at least some of the set of candidate venues based on the reduced respective confidence score;

selecting, by the server, a new highest ranked candidate venue as the visited venue of the venue visit using the adjusted ranking; and

providing, by the server and to a service associated with a user device, second one or more network communications comprising an indication of the selected visited venue that corresponds to the new highest ranked candidate venue, the indication causing content to be presented to the user on the user device based on the selected new highest ranked visited venue.

17. The computerized method of claim 16, wherein the service is an application on the user device and wherein the indication is received by the service on the user device.

18. The computerized method of claim 16, wherein the semantic information associated with the user is extracted from data generated corresponding to at least one of a search history generated at least partially by the user device, a text message sent or received by the user device, and an email sent or received by the user device.

19. The computerized method of claim 16, wherein at least some of the semantic information associated with the given candidate venue is inferred from portions of the sensor data generated by a plurality of user devices.