Systems and methods for the semantic modeling of advertising creatives in targeted search advertising campaigns

• US 9,972,030 B2
• Filed: 03/11/2013
• Issued: 05/15/2018
• Est. Priority Date: 03/11/2013
• Status: Active Grant

• Current Assignee: Criteo Technology SAS (Criteo SA)
• Original Assignee: Criteo SA
• Inventors: Zamanzadeh, Behzad, Zimmerman, Jr., Stephen John, Ramakrishnan, Cartic
• Primary Examiner(s): Reagan, James A
• Application Number:
  US13/792,718
• Publication Number:
  US 20140257967A1
• Time in Patent Office: 1,891 Days
• CPC Class Codes: G06Q 30/0256 User search

Abstract

Systems and methods for the semantic modeling of advertising creatives included in targeted search advertising campaigns in accordance with embodiments of the invention are disclosed. In one embodiment, an advertising creative generation device includes a processor, an advertising creative generation application, at least one semantic model and performance data, wherein the an advertising creative generation application configures the processor to obtain a set of existing advertising creatives, where at least one of the existing advertising creatives comprises an existing concept, identify at least one existing advertising strategy pattern, determine performance data for the at least one existing advertising strategy pattern based on the performance data, identify at least one performance pattern based on the performance data for the at least one existing advertising strategy pattern, and generate at least one recommended advertising strategy.

Claims

1. An advertising creative generation device, comprising:

a processor;

a memory readable by the processor and storing an advertising creative generation application;

semantic model storage storing semantic models, wherein a set of the semantic models comprises data describing concepts, advertising creatives, and semantic relationships between the concepts and the advertising creatives; and

performance data storage storing performance data for advertising creatives and a plurality of advertising strategies;

wherein the advertising creative generation application that when read by the processor directs the processor to;

obtain a set of existing advertising creatives for a set of associated concepts and a set of existing advertising strategies wherein each advertising strategy in the set of existing advertising strategies is based on the set of semantic models;

store the set of existing advertising creatives using the memory;

determine advertisement performance data describing the performance of the set of existing advertising creatives and the set of existing advertising strategies based on the performance data in the performance data storage associated with the set of existing advertising creatives and the set of semantic models upon which the advertising strategies in the set of existing advertising strategies is based;

store the advertisement performance data using the memory;

identify performance pattern data describing at least one performance pattern that describes a pattern between the performance of the set of existing advertising creatives and semantic relationships between the advertising creatives and the associated concepts;

store the performance pattern data using the memory;

generate at least one recommended advertising strategy for a particular concept based on the identified performance pattern data;

generate a new advertising creative for the particular concept for provision to a search engine provider based on the set of existing advertising creatives, the set of semantic models, and the at least one recommended advertising strategy; and

store the new advertising creative using the memory.

2. The system of claim 1, wherein the performance data for advertising creatives comprises campaign specific performance data for specific advertising creatives, associated keywords, and targeting data.

3. The system of claim 2, wherein the performance data for advertising creatives is selected from the group consisting of impressions, click-through rates, and conversion rates.

4. The system of claim 1, wherein generating the new advertising creative comprises modifying one of the advertising creatives from the set of existing advertising creatives based on the at least one recommended advertising strategy.

5. The system of claim 1, wherein the advertising creative generation application further directs the processor to transmit the new advertising creative to a device selected from the group consisting of an advertising server system and a search engine provider system.

6. The system of claim 1, wherein identifying performance pattern data comprises identifying at least one global performance pattern in global performance data, where the global performance data includes performance data for advertising creatives, keywords, and a plurality of advertising strategies across a plurality of existing targeted search advertising campaigns.

7. The system of claim 6, wherein identifying the at least one global performance pattern comprises performing a regression analysis based on the at least one existing advertising strategy and the global performance data.

8. The system of claim 6, wherein generating the at least one recommended advertising strategy comprises generating an advertising strategy that is not included in the set of existing advertising strategies and the global performance data.

9. The system of claim 1, wherein the at least one recommended advertising strategy is generated based on the identified performance pattern data, and a subset of the set of existing advertising creatives.

10. A method for generating advertising creatives, comprising:

obtaining a set of existing advertising creatives for a set of associated concepts and a set of existing advertising strategies using an advertising creative generation device wherein the advertising creative generation device comprises a processor and a memory, and the set of existing advertising creatives and the set of existing advertising strategies are stored using the advertising creative generation device and wherein each of the advertising strategies in the set of existing advertising strategies is based upon a set of sematic models comprising data describing concepts, advertising creatives, and semantic relationships between the concepts and the advertising creatives that is stored in the memory;

determining advertising performance data describing the performance of the set of existing advertising creatives and the set of existing advertising strategies based on performance data associated with the set of existing advertising creative and the set of semantic models upon which the set of advertising strategies is based;

storing the advertising performance data using the advertising creative generation device;

identifying performance pattern data describing at least one performance pattern that describes a pattern between the performance of the set of existing advertising creatives and semantic relationships between the advertising creatives and the associated concepts using the advertising creative generation device;

storing the performance pattern data using the advertising creative generation device;

generating at least one recommended advertising strategy for a particular concept based on the identified performance pattern data using the advertising creative generation device;

generating a new advertising creative for the particular concept to provide for use by a search engine provider based on the set of existing advertising creatives, the set of semantic models, and the at least one recommended advertising strategy using the advertising creative generation device; and

storing the new advertising creative using the advertising creative generation device.

11. The method of claim 10, wherein the performance data for the set of existing advertising creatives comprises campaign specific performance data for specific advertising creatives in the set of existing advertising creatives and associated keywords and targeting data.

12. The method of claim 11, wherein the performance data for the set of existing advertising creatives is selected from the group consisting of impressions, click-through rates, and conversion rates.

13. The method of claim 10, wherein generating the new advertising creative comprises modifying at least one of the existing advertising creatives based on the at least one recommended advertising strategy.

14. The method of claim 10, further comprising transmitting the new advertising creative to a device using the advertising creative generation device, where the device is selected from the group consisting of an advertising server system and a search engine provider system.

15. The method of claim 10, wherein identifying performance pattern data comprises identifying at least one global performance pattern in global performance data using the advertising creative generation device, where the global performance data includes performance data for advertising creatives, keywords, and a plurality of advertising strategies across a plurality of existing targeted search advertising campaigns.

16. The method of claim 15, wherein identifying the at least one global performance pattern comprises performing a regression analysis based on the at least one existing advertising strategy and the global performance data using the advertising creative generation device.

17. The method of claim 10, wherein generating the at least one recommended advertising strategy comprises generating an advertising strategy that is not included in the set of existing advertising strategies and the global performance data using the advertising creative generation device.

18. The method of claim 10, wherein the at least one recommended advertising strategy is generated based on the identified performance pattern data, and a subset of the set of existing advertising creatives using the advertising creative generation device.

19. The method of claim 10, wherein the semantic relationships between the advertising creatives and the associated concepts comprises a relationship between keywords used in the advertising creatives and attributes describing the associated concepts.

20. The method of claim 19, wherein generating the new advertising creative for the particular concept comprises identifying a new set of keywords to be used with the new advertising creative, wherein the new set of keywords have a similar relationship with the attributes of the particular concept as the semantic relationship between keywords used in the advertising creatives and attributes describing the associated concepts.