Context-based customization using semantic graph data

Patent metadata

Field	Value
Publication number	US-11829417-B2
Application number	US-201916678948-A
Country	US
Kind code	B2
Filing date	Nov 8, 2019
Priority date	Feb 5, 2019
Publication date	Nov 28, 2023
Grant date	Nov 28, 2023

Abstract

Methods, systems, and apparatus, including computer programs encoded on computer-readable storage media, for context-based customization using semantic graph data. In some implementations, semantic graph data indicating objects and relationships among the objects is stored. Usage data is generated to indicate (i) levels of usage of the objects and (ii) contexts in which the objects are used. A query is received that is associated with a user and that includes data indicating a context of the user. Customized weights for connections among the objects is determined based on (i) the levels of usage indicated by the usage data and (ii) scores indicating a degree of similarity among the contexts indicated by the usage data and the context of the user. A response to the query is provided based on the customized weights for the connections among the objects indicated by the semantic graph data.

First claim

What is claimed is:

1. A method performed by one or more computers, the method comprising:
storing, by the one or more computers, semantic graph data indicating objects and relationships among the objects;
generating, by the one or more computers, usage data indicating (i) instances of usage of the objects by multiple users and (ii) contexts in which the objects have been used by the multiple users;
receiving, by the one or more computers, (i) a request associated with a user and (ii) context data indicating a context of the user associated with the request;
in response to receiving the request, determining, by the one or more computers, customized weights for connections among the objects, wherein the customized weights are customized for the user and are based on accesses of the multiple users to the objects, wherein the customized weights are determined by:
determining similarity scores each indicating a degree of similarity between a different one of the contexts indicated by the usage data and the context associated with the request; and
generating each of the customized weights based on a different weighted combination of usage measures for different contexts, wherein instances of usage by the multiple users in the different contexts are given different levels of influence in the weighted combination according to the corresponding similarity scores indicating similarity between the contexts in which the instances of usage occurred and the context of the user; and

providing, by the one or more computers, a response to the request that is determined based on the customized weights for the connections among the objects indicated by the semantic graph data.

2. The method of claim 1 , wherein the request is a query, and wherein the response to the request comprises one or more search results responsive to the query, wherein the one or more search results are selected or ranked based on the customized weights for the connections.

3. The method of claim 1 , wherein the request is a request for at least one of a recommendation, a suggestion, or a prediction determined using a semantic graph.

4. The method of claim 1 , wherein the response comprises at least one of a search result, a metric, an attribute, an entity name, a data source, a document, a dashboard, or a visualization.

5. The method of claim 1 , wherein the usage data indicates multiple accesses to a first object associated with different contexts, and wherein determining the customized weights comprises adjusting weights of connections to the first object according to a similarity of the different contexts to the context associated with the request.

6. The method of claim 1 , wherein determining the customized weights comprises:
determining an adjustment for a first weight indicated by the semantic graph data for a connection between a first object and a second object, the adjustment being based on one or more similarity scores indicating a degree of similarity between a context for a previous access to the first object and the context associated with the request; and
applying the adjustment to the first weight to determine a customized weight for the connection between the first object and the second object.

7. The method of claim 6 , wherein the adjustment is a scaling factor or an additive offset.

8. The method of claim 1 , wherein determining the customized weights comprises:
determining a subset of the usage data indicating use of a first object associated with a context having a similarity with the context associated with the request that satisfies a threshold; and
determining a customized weight for a connection with the first object based on the subset of the usage data.

9. The method of claim 1 , wherein determining the customized weights comprises determining a customized weight for a connection between a first object and a second object based on a weighted combination of uses of the first object, wherein the uses of the first object are weighted according to the respective similarity of the contexts of the uses and the context associated with the request.

10. The method of claim 1 , wherein determining the customized weights comprises determining a customized weight for a connection between a first object and a second object based on a weighted combination of different usage measures of the first object, the different usage measures each representing different groups of uses of the first object.

11. The method of claim 10 , wherein the different groups of uses comprise groups of multiple instances of use of the first object, the groups respectively representing uses associated with different users, different user roles, different organizations, different departments, different locations, different time ranges, different types of access, and/or different devices.

12. The method of claim 1 , wherein determining the customized weights comprises determining a customized weight for a connection of a first object and a second object that includes multiple adjustments based on multiple different contextual factors.

13. The method of claim 1 , wherein the context associated with the request and the contexts associated with the usage data each indicate at least one of a a time, a location, a topic, a keyword, a task, an application, a type of access, a type of device of device, or a mode of access.

14. The method of claim 1 , wherein the context associated with the request indicates characteristics of a logical or physical environment of a device that submitted the request.

15. The method of claim 1 , wherein generating the usage data comprises tracking multiple types of actions involving the objects, the actions including at least one of requesting, serving, reading, writing, saving, printing, sharing, or selecting the objects.

16. The method of claim 1 , wherein determining the customized weights for the connections among the objects comprises determining user-specific weights for the connections, the user-specific weights being based at least in part on (i) prior actions of the user with respect to the objects and/or (ii) aggregations of usage data of multiple users determined based on characteristics of the user or prior actions of the user.

17. The method of claim 1 , wherein the semantic graph data indicates connections among the objects that represent different types of relationships among the objects, wherein the different types of relationships include at least one of a dependency of an object on another object, a co-occurrence of an object with another object, an object being an instance of a class or category represented by another object, an object being a part of another object, or an object using or accessing another object.

18. A system comprising:
one or more computers; and
one or more non-transitory computer-readable media storing instructions that, when executed by the one or more computers, cause the one or more computers to perform operations comprising:
storing, by the one or more computers, semantic graph data indicating objects and relationships among the objects;
generating, by the one or more computers, usage data indicating (i) instances of usage of the objects by multiple users and (ii) contexts in which the objects have been used by the multiple users;
receiving, by the one or more computers, (i) a request associated with a user and (ii) context data indicating a context of the user associated with the request;
in response to receiving the request, determining, by the one or more computers, customized weights for connections among the objects, wherein the custom

Assignees

• Microstrategy Inc

Inventors

• Abhyankar Saurabh
• Rigney Scott
• Lang Timothy

Classifications

• G06F16/9024 Primary

  Graphs; Linked lists (G06F16/9027 takes precedence) · CPC title

• G06F16/24575

  using context · CPC title

• G06F16/24578

  using ranking · CPC title

• G06F16/9035

  Filtering based on additional data, e.g. user or group profiles · CPC title

• G06F16/9038

  Presentation of query results · CPC title