System and method for providing customized advertising on the World Wide Web

• US 6,654,725 B1
• Filed: 11/09/1999
• Issued: 11/25/2003
• Est. Priority Date: 11/09/1998
• Status: Expired due to Fees

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Langheinrich, Marc, Nakamura, Atsuyoshi
• Primary Examiner(s): Carlson, Jeffrey D.
• Application Number:
  US09/436,241
• Time in Patent Office: 1,477 Days
• Field of Search: 705/1, 705/10, 705/14, 705/26
• US Class Current: 705/14.52
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0254 based on statistics
• CPC Class Codes: G06Q 30/0256 User search
• CPC Class Codes: G06Q 30/0277 Online advertisement

Abstract

A system and method for customized advertisement selection and delivery on the World Wide Web (WWW) upon the Internet. The advertising system has a database server which stores advertisements and their campaign information, and an advertisement server which generates electronic advertisements available to a client system. In the system, a customization process which customized the electronic advertisements to be delivered to each client system is performed. A user connects to a web site and is presented with an editorial page or a list of search results. The system inserts a customized advertisement into the page that matches the page content or search topic. No identifiable data is collected during the interaction with the user. Advertisers can specify display constraints for each advertisement. The system will adapt all unrestricted parameters in order to maximize the user'"'"'s click-through probability.

Claims

1. A method of providing an electronic advertisement from a plurality of electronic advertisements to a client system, the method comprising the steps of:

decoding one or more customization parameters embedded in a request from the client system;

querying a database for a list of display probabilities for relevant values associated with the one or more customization parameters;

computing an overall display probability for the request based on a result of the querying step;

selecting one of said plurality of electronic advertisements according to the overall display probability; and

wherein said list of display probabilities is generated by calculating display probabilities for one or more customization parameters based on an associated number of click-throughs, from a plurality of users with respect to each of said plurality of electronic advertisements.

2. The method according to claim 1 wherein the list of display probabilities queried in the querying step are pre-computed by a learning system.

3. The method according to claim 1 wherein the relevant values associated with the customization parameters are stored in the database.

4. The method according to claim 1 further comprising the step of:

recording display and click-through statistics in the database for each set of customization parameters decoded in the decoding step with respect to the one electronic advertisement selected in the selecting step.

5. The method according to claim 1 wherein the querying step further comprises the steps of:

providing a learning system which pre-computes the display probabilities; and

employing said learning system to periodically update the display probabilities.

6. The method according to claim 5 further comprising the steps of:

obtaining constraint data from the database for use by the learning system in updating the display probabilities;

obtaining performance data from the database for use by the learning system in updating the display probabilities; and

obtaining one or more customization parameter distributions from the database for use by the learning system in updating the display probabilities.

7. The method according to claim 5, further comprising the step of:

storing a new list of display probabilities in the database based on a result of said learning system updating said display probabilities.

8. The method according to claim 7 wherein the new list of display probabilities stored to the database is calculated so as to maximize a total click-through rate with respect to

(a) an estimated customization parameter distribution, and

(b) an estimated click-through probability for one or more customization parameters and each of said plurality of electronic advertisements.

9. The method according to claim 8 wherein said total click-through rate is further maximized according to an associated constraint of a daily impression rate for each of said plurality of electronic advertisements.

10. The method according to claim 9 wherein said total click-through rate is further maximized under a custom display constraint set by an advertiser.

11. The method according to claim 1, wherein the display probabilities are calculated using linear programming.

12. The method according to claim 1, wherein the display probabilities are calculated using linear programming under one or more advertising campaign constraints.

13. An apparatus which provides electronic advertisements to a client system coupled to the apparatus, the apparatus comprising:

a database which stores a plurality of electronic advertisements and corresponding campaign information;

an advertisement server accessing said database for making one of said plurality of electronic advertisements available to the client system; and

customization means connected to said advertisement server for performing a customization process which customizes the one electronic advertisement to be delivered to the client system; and

wherein said customization means comprises;

means for calculating display probabilities for one or more customization parameters based on a number of click-throughs from a plurality users with respect to each of said plurality of electronic advertisement; and

means for selecting the one electronic advertisement to be delivered based on the calculated display probabilities, the campaign information corresponding to said plurality of electronic advertisements, and one or more customization parameters received from the client system.

14. The apparatus according to claim 13 wherein the database stores a plurality of electronic advertisement titles, links and banner graphics, associated with said plurality of electronic advertisements.

15. The apparatus according to claim 13 wherein the database stores a plurality of advertisement campaigns associated with electronic advertisements.

16. The apparatus according to claim 13 wherein the database stores a list of customization parameters as a part of said corresponding campaign information.

17. The apparatus according to claim 16 wherein the database stores display statistics for each of said plurality of electronic advertisements, and wherein the database stores valid combinations of customization parameters from said list of customization parameters and associated values.

18. The apparatus according to claim 17 wherein the database stores click-through statistics for each of said plurality of electronic advertisements and for each of said valid combinations of customization parameters and associated values.

19. The apparatus according to claim 13, wherein the display probabilities are calculated using linear programming.

20. The apparatus according to claim 13, wherein the display probabilities are calculated using linear programming under one or more advertising campaign constraints.