Content selection

Patent metadata

Field	Value
Publication number	US-9767196-B1
Application number	US-201314085153-A
Country	US
Kind code	B1
Filing date	Nov 20, 2013
Priority date	Nov 20, 2013
Publication date	Sep 19, 2017
Grant date	Sep 19, 2017

Abstract

Systems and methods of the present disclosure are directed generally to facilitating content selection by identifying low impact criteria. In some implementations, a data processing system accesses a data structure storing, in a memory element, a plurality of impression records. Each impression record can include one or more features and an indication of user interest corresponding to a content impression. The data processing system can identify a combination feature based on at least two of the features. The data processing system can execute a statistical model (e.g., logistic regression model) using the impression records and the combination feature. The data processing system can determine a weight for the combination feature. Responsive to the weight being less than a threshold, the data processing system can transmit an indication to disable the combination feature for selecting content associated with the plurality of impression records.

First claim

What is claimed is:

1. A method of selecting content via a computer network, comprising:
accessing, by one or more processors, a data structure storing, in memory, a plurality of impression records corresponding to a plurality of content impressions of a content item, each of the plurality impression records including a plurality of features and an indication of interest in the content item, the plurality of features including at least two of a web page category, a content category, a geographic region, and a language;
identifying, from an impression record of the plurality of impression records, a combination feature that includes a combination of at least two of the web page category, the content category, the geographic region, and the language;
identifying, via the one or more processors, a statistical model using the plurality of impression records and the combination feature;
determining, via the statistical model, a weight for the combination feature, the weight indicating interest in content impressions associated with the combination feature; and
preventing, by the one or more processors responsive to the weight being less than a threshold, a content selector executing on one or more servers from selecting the content item for display in the presence of the combination feature that includes at least two features of the plurality of features.

2. The method of claim 1 , wherein the statistical model comprises a logistic regression model.

3. The method of claim 1 , wherein the plurality of impression records correspond to a type of content provider.

4. The method of claim 1 , wherein the indication of interest indicates receiving a selection of a content item of the content impression or not receiving a selection of the content item of the content impression.

5. The method of claim 1 , comprising:
determining, via the statistical model, a weight for each of the plurality of features; and
identifying a lowest weight based on the weight for each of the plurality of features; and
transmitting an indication to disable a feature corresponding to the lowest weight.

6. The method of claim 1 , comprising:
modifying the threshold using content performance feedback, the content performance feedback including at least one of a click through rate and conversion rate.

7. The method of claim 1 , comprising:
receiving, via the computer network, a request for content corresponding to the combination feature;
identifying that the combination feature is disabled for selecting the content item associated with the plurality of impression records;
identifying that the combination feature is enabled for a second content item associated with a second plurality of impression records; and
selecting the second content item as a candidate for display on a computing device associated with the request.

8. The method of claim 1 , comprising:
determining, via the statistical model, based on a second plurality of impression records, a second weight for the combination feature, the second plurality of impression records corresponding to a different time interval; and
transmitting to the content selector, responsive to the second weight being greater than a second threshold, an indication to enable the combination feature for selecting content associated with the second plurality of impression records.

9. A system for facilitating content selection via a computer network, comprising:
one or more data processing systems comprising memory and one or more processors configured to:
access a data structure storing, in the memory, a plurality of impression records corresponding to a plurality of content impressions of a content item, each of the plurality impression records including a plurality of features and an indication of interest in the content item, the plurality of features including at least two of a web page category, a content category, a geographic region, and a language;
identify, from an impression record of the plurality of impression records, a combination feature that includes a combination of at least two of the web page category, the content category, the geographic region, and the language;
identify a statistical model using the plurality of impression records and the combination feature;
determine, via the statistical model, a weight for the combination feature, the weight indicating interest in content impressions associated with the combination feature; and
prevent, via the computer network responsive to the weight being less than a threshold, a content selector executing on one or more servers from selecting the content item for display in the presence of the combination feature that includes at least two features of the plurality of features.

10. The system of claim 9 , wherein the statistical model comprises a logistic regression model.

11. The system of claim 9 , wherein the plurality of impression records correspond to a type of content provider.

12. The system of claim 9 , wherein the data processing system is further configured to perform operations comprising:
determining, via the statistical model, a weight for each of the plurality of features; and
identifying a lowest weight based on the weight for each of the plurality of features; and
providing an indication to disable a feature corresponding to the lowest weight.

13. The system of claim 9 , wherein the data processing system is further configured to perform operations comprising:
modifying the threshold using content performance feedback, the content performance feedback including at least one of a click through rate and conversion rate.

14. The system of claim 9 , wherein the data processing system is further configured to perform operations comprising:
receiving, via the computer network, a request for content corresponding to the combination feature;
identifying that the combination feature is disabled for selecting content associated with the plurality of impression records;
identifying that the combination feature is enabled for second content associated with a second plurality of impression records; and
selecting the second content as a candidate for display on a computing device associated with the request.

15. The system of claim 9 , wherein the data processing system is further configured to perform operations comprising:
determining, via the statistical model, a second weight for a feature of the plurality of features; and
providing, responsive to the second weight being greater than a second threshold, an indication to enable the feature for selecting content associated with the plurality of impression records.

16. A computer-readable storage device comprising processor executable instructions for facilitating content selection via a computer network, the instructions comprising instructions to:
access a data structure storing, in memory, a plurality of impression records corresponding to a plurality of content impressions of a content item, each of the plurality impression records including a plurality of features and an indication of interest in the content item, the plurality of features including at least two of a web page category, a content category, a geographic region, and a language;
identify, from an impression record of the plurality of impression records, a combination feature that includes a combination of at least two of the web page category, the content category, the geographic region, and the language;
identify a statistical model using the plurality of impression records and the combination feature;
determine, via the statistical model, a weight for the combination fe

Assignees

• Google Inc

Inventors

• Peng Yefei
• Harris Christopher Kenneth

Classifications

• G06F16/95 Primary

  Retrieval from the web · CPC title

• G06F17/30861 Primary

  Physics · mapped topic

• G06F16/9535 Primary

  Search customisation based on user profiles and personalisation · CPC title