Method and apparatus for providing web advertisements to users

Patent metadata

Field	Value
Publication number	US-11734728-B2
Application number	US-202016795315-A
Country	US
Kind code	B2
Filing date	Feb 19, 2020
Priority date	Feb 20, 2019
Publication date	Aug 22, 2023
Grant date	Aug 22, 2023

Abstract

A method and apparatus for providing Web advertisements to online users is disclosed. A balanced set of negative data points and positive data points is derived from a log of Ad impressions and used to train a classifier. In response to an Ad request signal, a plurality of Ads is retrieved from a database. The Ad request signal indicates a request to provide an Ad for a slot available on a Web page associated with a website. The signal is provided in relation to an access of the Web page by an online user and includes information related to the online user. A choice of an Ad is predicted based on the information related to the online user and the plurality of Ads. The Ad is provided to a Web server to cause display of the Ad on the slot when the Web page is displayed to the online user.

First claim

The invention claimed is:

1. A computer-implemented method for providing Web advertisements to online users, the method comprising:
training a classifier of an apparatus associated with an advertisement (Ad) server by using a balanced set of negative data points and positive data points derived from a log of Ad impressions, the balanced set configured based at least in part on identifying a consensus among base cluster representations generated by clustering the negative data points using two or more clustering algorithms;
receiving, by the apparatus, a signal indicating a request to provide at least one Ad for a slot available on a Web page associated with a website, the signal provided to the apparatus in relation to an access request for the Web page by an online user, the signal at least in part comprising information related to the online user;
retrieving a plurality of advertisements (Ads), by the apparatus, from a database associated with the apparatus in response to the receipt of the signal;
predicting, by the apparatus using the classifier, a choice of an Ad from among the plurality of Ads based at least in part on the information related to the online user and the plurality of Ads; and
providing, by the apparatus, the Ad to a Web server associated with the website to cause display of the Ad on the slot available on the Web page when the Web page is displayed to the online user.

2. The method of claim 1 , wherein the log of Ad impressions comprises information related to at least one of:
Ads impressions associated with one or more Ad publishers, and
online users of the one or more Ad publishers served with the Ad impressions.

3. The method of claim 2 , further comprising:
extracting, by the apparatus, features from the information related to the Ad impressions and the online users, wherein extracting features from the information related to the Ad impressions at least comprises extracting information related to baseline content, theme, background, message and call-to-action button from a respective Ad; and
tagging, by the apparatus, each feature with a label indicative of whether a click or a conversion is associated with the respective Ad, wherein the tagging of each feature is configured to generate the negative data points and the positive data points.

4. The method of claim 3 , further comprising:
generating, by the apparatus, an ensemble cluster based on identifying the consensus among the base cluster representations of the negative data points generated using the two or more clustering algorithms.

5. The method of claim 4 , further comprising:
selecting, by the apparatus, a plurality of data points within a predefined distance from a cluster center of the ensemble cluster, wherein the selected plurality of data points configure a first data of data points, the first set of data points representative of an undersampled form of the negative data points.

6. The method of claim 5 , further comprising:
generating, by the apparatus, a second set of data points by oversampling the positive data points, wherein the first set of data points and the second set of data points configure the balanced set of the negative data points and the positive data points.

7. The method of claim 6 , wherein the two or more clustering algorithms used to cluster the negative data points to generate the base cluster representations comprise at least one of Latent Dirichlet Allocation (LDA) clustering algorithm, Gaussian Mixture Models (GMM) clustering algorithm and K-Means clustering algorithm, and wherein Synthetic Minority Over-sampling Technique (SMOTE) algorithm is used for oversampling the positive data points.

8. The method of claim 1 , further comprising:
receiving, by the apparatus, information related to Ad impressions from one or more demand-side platforms (DSPs) and one or more Ad Exchanges, wherein the log of Ad impressions is generated based on the information related to Ad impressions received from the one or more DSPs and the one or more Ad Exchanges; and
assigning relevancy weights, by the apparatus, to the Ad impressions based on prior knowledge of quality of traffic associated with respective DSP from among the one or more DSPs and respective Ad Exchange from among the one or more Ad Exchanges.

9. The method of claim 1 , wherein the classifier is trained to meet a predefined objective, the predefined objective corresponding to at least one objective from among an objective to increase a likelihood of the online user clicking on the Ad, an objective to increase a likelihood of the online user engaging in a purchase transaction in relation to the Ad, and an objective to increase an awareness of the online user.

10. An apparatus for providing Web advertisements to online users, the apparatus comprising:
a memory for storing instructions; and
a processor configured to execute the instructions and thereby cause the apparatus to at least:
train a classifier by using a balanced set of negative data points and positive data points derived from a log of Ad impressions, the balanced set configured based at least in part on identifying a consensus among base cluster representations generated by clustering the negative data points using two or more clustering algorithms;
receive a signal indicating a request to provide at least one Ad for a slot available on a Web page associated with a website, the signal provided to the apparatus in relation to an access request for the Web page by an online user, the signal at least in part comprising information related to the online user;
retrieve a plurality of advertisements (Ads), by the apparatus, from a database associated with the apparatus in response to the receipt of the signal;
predict using the classifier, a choice of an Ad from among the plurality of Ads based at least in part on the information related to the online user and the plurality of Ads; and
provide the Ad to a Web server associated with the website to cause display of the Ad on the slot available on the Web page when the Web page is displayed to the online user.

11. The apparatus of claim 10 , wherein the log of Ad impressions comprises information related to at least one of:
Ads impressions associated with one or more Ad publishers, and
online users of the one or more Ad publishers served with the Ad impressions.

12. The apparatus of claim 11 , wherein the apparatus is further caused to:
extract features from the information related to the Ad impressions and the online users, wherein extracting features from the information related to the Ad impressions at least comprises extracting information related to baseline content, theme, background, message and call-to-action button from a respective Ad; and
tag each feature with a label indicative of whether a click or a conversion is associated with the respective Ad, wherein the tagging of each feature is configured to generate the negative data points and the positive data points.

13. The apparatus of claim 12 , wherein the apparatus is further caused to:
generate an ensemble cluster based on identifying the consensus among the base cluster representations of the negative data points generated using the two or more clustering algorithms.

14. The apparatus of claim 13 , wherein the apparatus is further caused to:
select a plurality of data points within a predefined distance from a cluster center of the ensemble cluster, wherein the selected plurality of data points configure a first data of data points, the first set of data points representative of an undersampled form of the negative data points.

15. The apparatus of claim 14

Assignees

• [24]7 Ai Inc

Inventors

• Gupta Nitin
• Kumar Gaurav

Classifications

• G06Q30/0277 Primary

  Online advertisement · CPC title

• G06F18/23213

  with fixed number of clusters, e.g. K-means clustering · CPC title

• G06Q30/0246

  Traffic · CPC title

• G06Q30/0271

  Personalized advertisement · CPC title

• G06F18/2321

  using statistics or function optimisation, e.g. modelling of probability density functions · CPC title