PROBABILISTIC FRAMEWORK FOR DETERMINING DEVICE ASSOCIATIONS

• US 20200134497A1
• Filed: 10/26/2018
• Published: 04/30/2020
• Est. Priority Date: 10/26/2018
• Status: Active Grant

• Current Assignee: Salesforce.com, Inc.
• Original Assignee: Salesforce.com, Inc.
• Inventors: Salomon, Yacov, Budd, Jonathan
• Granted Patent: US 11,030,545 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 16/248 Presentation of query results
• CPC Class Codes: G06F 16/9024 Graphs; Linked lists G06F16/9027 takes precedence
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks

Abstract

Methods, systems, and devices for determining device associations are described. Some database systems may store information related to device characteristics. Each of these devices may be operated by one or more users, and each user may operate one or more devices. In some cases, information about users may be more valuable than information about devices. As such, a system may determine probable associations between devices, where an association can correspond to operation by a same user. To determine device associations, the system may perform a machine-learning process (e.g., using probabilistic soft logic (PSL) and a hinge-loss Markov Random Field (HL-MRF) model) on input device characteristics and connection information to generate a probability density function. The probability density function may indicate associations between devices within the system. Based on one or more thresholds, the system may determine sets of associated devices and may transmit this association information for analysis or display.

Claims

1. A method for device association, comprising:

receiving, from a database, a plurality of device characteristics corresponding to a plurality of devices, wherein each device of the plurality of devices is associated with a respective set of device characteristics of the plurality of device characteristics;

receiving connection information for the plurality of devices;

performing a machine-learning process based at least in part on the plurality of device characteristics and the connection information;

determining a probability density function for device association based at least in part on an output of the machine-learning process;

identifying a plurality of sets of associated devices based at least in part on the probability density function for device association; and

transmitting, to a device, information for display corresponding to at least one set of the identified plurality of sets of associated devices.

2. The method of claim 1, wherein performing the machine-learning process comprises:

generating a hinge-loss Markov Random Field (HL-MRF) model, wherein the probability density function is determined based at least in part on the HL-MRF model.

3. The method of claim 2, wherein generating the HL-MRF model comprises:

using probabilistic soft logic (PSL) to generate the HL-MRF model.

4. The method of claim 1, further comprising:

receiving a weight threshold for association, wherein the plurality of sets of associated devices is determined based at least in part on the probability density function for device association and the weight threshold.

5. The method of claim 1, further comprising:

receiving a constraint for a constraint optimization process, wherein the plurality of sets of associated devices is determined based at least in part on the probability density function for device association and the constraint for the constraint optimization process.

6. The method of claim 5, wherein the constraint for the constraint optimization process comprises a threshold number of total device associations, a threshold number of associations per device, a threshold number of devices per association, or a combination thereof.

7. The method of claim 1, further comprising:

generating, based at least in part on the probability density function for device association, a device graph comprising a plurality of nodes and a plurality of weighted edges, wherein each node of the plurality of nodes corresponds to a device of the plurality of devices and each weighted edge of the plurality of weighted edges connects two nodes of the plurality of nodes and corresponds to a probability that two devices corresponding to the connected two nodes are associated.

8. The method of claim 7, further comprising:

removing a set of weighted edges from the device graph based at least in part on one or more association constraints; and

identifying a plurality of device sub-graphs resulting from the removing the set of weighted edges, wherein connected nodes of the plurality of device sub-graphs indicate the identified plurality of sets of associated devices.

9. The method of claim 1, further comprising:

iteratively updating weighting values for the probability density function for device association based at least in part on updating the plurality of device characteristics, updating the plurality of devices, updating the connection information, or a combination thereof.

10. The method of claim 1, further comprising:

receiving, from the device, a query message indicating a first device of the plurality of devices; and

determining the at least one set of the identified plurality of sets of associated devices based at least in part on the indicated first device, wherein the at least one set of the identified plurality of sets of associated devices comprises the indicated first device and one or more additional devices, and wherein transmitting the information for display comprises;

transmitting, to the device in response to the query message, an indication of the one or more additional devices.

11. The method of claim 1, wherein each set of the plurality of sets of associated devices is associated with a respective entity.

12. The method of claim 11, further comprising:

receiving, from the device, a query message indicating an entity; and

determining the at least one set of the identified plurality of sets of associated devices based at least in part on the indicated entity, wherein the at least one set of the identified plurality of sets of associated devices is associated with the indicated entity, and wherein transmitting the information for display comprises;

transmitting, to the device in response to the query message, an indication of each associated device in the at least one set of the identified plurality of sets of associated devices.

13. The method of claim 11, wherein the entity comprises a single user, a set of users, a household, an organization, or a combination thereof.

14. The method of claim 1, wherein the plurality of device characteristics comprises authentication activity information, user agent information, device types, internet protocol (IP) addresses, geographic location information, web activity information, activity timestamps, or a combination thereof.

15. The method of claim 1, wherein the plurality of devices comprises mobile devices, laptops, desktops, internet of things (IoT) devices, internet of everything (IoE) devices, web browsers, web cookies, or a combination thereof.

16. An apparatus for device association, comprising:

a processor;

memory in electronic communication with the processor; and

instructions stored in the memory and executable by the processor to cause the apparatus to;

receive, from a database, a plurality of device characteristics corresponding to a plurality of devices, wherein each device of the plurality of devices is associated with a respective set of device characteristics of the plurality of device characteristics;

receive connection information for the plurality of devices;

perform a machine-learning process based at least in part on the plurality of device characteristics and the connection information;

determine a probability density function for device association based at least in part on an output of the machine-learning process;

identify a plurality of sets of associated devices based at least in part on the probability density function for device association; and

transmit, to a device, information for display corresponding to at least one set of the identified plurality of sets of associated devices.

17. The apparatus of claim 16, wherein the instructions to perform the machine-learning process are executable by the processor to cause the apparatus to:

generate a hinge-loss Markov Random Field (HL-MRF) model, wherein the probability density function is determined based at least in part on the HL-MRF model.

18. The apparatus of claim 16, wherein the instructions are further executable by the processor to cause the apparatus to:

receive a weight threshold for association, wherein the plurality of sets of associated devices is determined based at least in part on the probability density function for device association and the weight threshold.

19. The apparatus of claim 16, wherein the instructions are further executable by the processor to cause the apparatus to:

receive a constraint for a constraint optimization process, wherein the plurality of sets of associated devices is determined based at least in part on the probability density function for device association and the constraint for the constraint optimization process.

20. A non-transitory computer-readable medium storing code for device association, the code comprising instructions executable by a processor to:

receive, from a database, a plurality of device characteristics corresponding to a plurality of devices, wherein each device of the plurality of devices is associated with a respective set of device characteristics of the plurality of device characteristics;

receive connection information for the plurality of devices;

perform a machine-learning process based at least in part on the plurality of device characteristics and the connection information;

determine a probability density function for device association based at least in part on an output of the machine-learning process;

identify a plurality of sets of associated devices based at least in part on the probability density function for device association; and

transmit, to a device, information for display corresponding to at least one set of the identified plurality of sets of associated devices.