Analytics system entity resolution

Patent metadata

Field	Value
Publication number	US-11550859-B2
Application number	US-201916569484-A
Country	US
Kind code	B2
Filing date	Sep 12, 2019
Priority date	Sep 12, 2019
Publication date	Jan 10, 2023
Grant date	Jan 10, 2023

Abstract

Techniques and systems are described for analytics system entity resolution. Typed higher-order node combinations are determined within a dataset, and an amount of similarity between two arbitrary nodes within the dataset is determined based on the typed higher-order node combinations. The amount of similarity enables the digital analytics to accurately perform source resolution of portions of the dataset to a respective source, and may be utilized to control output of digital content to a client device.

First claim

What is claimed is:

1. A method implemented by at least one computing device, the method comprising:
receiving, by the at least one computing device, a connectivity pattern that defines a typed four cycle including a plurality of nodes having two center nodes and at least two additional nodes, each of the two center nodes linked via respective edges to both of the at least two additional nodes, two center nodes associated with a first node category of a source identifier and the at least two additional nodes associated with at least a second node category;
receiving, by the at least one computing device, a dataset comprising a node combination having:
nodes that include a first source identifier associated with a first session and a second source identifier associated with a second session;
node connections corresponding to at least the node connection of the connectivity pattern; and
node categories corresponding to the first node category and the second node category other than the first node category of the connectivity pattern;

determining that the node combination corresponds to the connectivity pattern;
determining a probability that the first source identifier associated with the first session and the second source identifier associated with the second session correspond to a single entity based on the node combination corresponding to the connectivity pattern;
resolving, by the at least one computing device, the first session associated with the first source identifier and the second session associated with the second source identifier as corresponding to the single entity based on the probability; and
controlling, by the at least one computing device, output of digital content to the single entity based on the first session and the second session.

2. The method of claim 1 , wherein determining that the connectivity pattern corresponds to the node combination further includes determining the connectivity pattern corresponds to a plurality of node combinations from the dataset.

3. The method of claim 1 , further comprising determining another connectivity pattern corresponds to another node combination from the dataset.

4. The method of claim 1 , wherein the second node category is associated with a web page, a location, an IP address, or a source agent.

5. The method of claim 1 , wherein the first session and the second session each include a plurality of nodes associated with the second source identifier, wherein the second source identifier corresponds to a web page, a location, an IP address, or a source agent.

6. The method of claim 1 , wherein the resolving includes determining an amount of similarity between nodes associated with the first and second source identifiers.

7. The method of claim 6 , wherein the amount of similarity is based on a first weighting factor associated with the first category and a second weighting factor associated with the second category.

8. The method of claim 1 , wherein the connectivity pattern is a typed four-cycle in which each of the at least two additional nodes is linked via the respective edges to both of the center nodes.

9. The method of claim 8 , wherein the typed four-cycle includes two center nodes of a first category and the at least two additional nodes each of the second category.

10. The method of claim 8 , wherein the typed four-cycle includes two center nodes of the first category, a first said additional node of the second category, and a second said additional node of a third category.

11. A system comprising one or more processors and one or more computer-readable media that, when executed by the one or more processors, are configured to implement:
a heterogeneous network representation module configured to receive a dataset comprising a node combination having: a first session associated with a first source identifier and a second session associated with a second source identifier and one or more node connections;
a higher-order node combination module configured to:
receive a connectivity pattern that defines a typed four cycle including a plurality of nodes having two center nodes and at least two additional nodes, each of the two center nodes linked via respective edges to both of the at least two additional nodes; and
determine the connectivity pattern corresponds to the node combination from the dataset; and

a linking module configured to:
determine a probability that the first source identifier associated with the first session and the second source identifier associated with the second session correspond to a single entity based on the node combination corresponding to the connectivity pattern; and
resolve the first session associated with the first source identifier and the second session associated with the second source identifier as corresponding to the single entity based on the probability.

12. The system of claim 11 , further comprising a digital content control module configured to control output of digital content to the single entity based on the first and second sessions.

13. The system of claim 11 , wherein determining that the connectivity pattern corresponds to the node combination from the dataset includes determining the connectivity pattern corresponds to a plurality of node combinations from the dataset.

14. The system of claim 11 , wherein the resolving includes determining an amount of similarity between nodes associated with the first and second source identifiers.

15. The system of claim 11 , wherein the connectivity pattern is a typed four-cycle in which the two center nodes corresponding to a first category and the at least two additional nodes corresponding to a second category.

16. The system of claim 11 , wherein the connectivity pattern is a typed four-cycle in which the two center nodes correspond to a first category and the at least two additional nodes correspond to at least a second category and a third category.

17. A system comprising:
means for receiving a connectivity pattern defining a typed four cycle including a plurality of nodes having two center nodes and two additional nodes, each of the two additional nodes linked via respective edges to both of the center nodes, with the two center nodes including a first node category of a client device, and the two additional nodes associated with at least a second node category other than a client device;
means for receiving a dataset comprising a node combination having:
nodes that include a first source identifier associated with a first session and a second source identifier associated with a second session;
node connections corresponding to at least the node connection of the connectivity pattern; and
node categories corresponding to the first node category of the client device and the second node other than the client device category of the connectivity pattern;

means for determining that the node combination corresponds to the connectivity pattern;
means for determining a probability that the first source identifier associated with the first session and the second source identifier associated with the second session correspond to a single entity based on the node combination corresponding to the connectivity pattern;
means for resolving the first session associated with the first source identifier and the second session associated with the second source identifier as corresponding to the single entity based on the probability; and
means for controlling output of digital content to the single entity based on the first session and the second session.

18. The system of cl

Assignees

• Adobe Inc

Inventors

• Rossi Ryan A
• Kim Sungchul
• Koh Eunyee
• Rao Anup Bandigadi
• Stringham Russell R

Classifications

• G06F16/9535 Primary

  Search customisation based on user profiles and personalisation · CPC title

• G06F16/955 Primary

  using information identifiers, e.g. uniform resource locators [URL] · CPC title