System and method for integrated recommendations

• US 7,966,219 B1
• Filed: 09/24/2004
• Issued: 06/21/2011
• Est. Priority Date: 09/24/2004
• Status: Active Grant

• Current Assignee: Versata Development Group, Inc. (ESW Capital, LLC)
• Original Assignee: Versata Development Group, Inc. (ESW Capital, LLC)
• Inventors: Wilbur, Thomas W., Thier, Neil C., Hyams, Chris, Singh, James
• Primary Examiner(s): Airapetian; Mila
• Application Number:
  US10/950,354
• Time in Patent Office: 2,461 Days
• Field of Search: 705/26, 705/27, 705/14
• US Class Current: 705/26.8
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0631 Item recommendations
• CPC Class Codes: G06Q 30/0633 Lists, e.g. purchase orders, compilation or processing

Abstract

A recommendation appliance, system and method are provided for generating and deploying additional web page content or functionality (e.g., retail recommendations) to an existing web page server system. For example, the present invention may be embodied as a reverse proxy server that is inserted as an intermediate network node between a web server and the end users accessing the web server. In this position, the recommendation appliance can introduce recommendation messages to web pages generated by the web server without requiring any modification to the code or architecture of the web server. In addition, the appliance may separately track the transaction activities of end users who receive recommendation messages and the transaction activities of end users who do not receive recommendation messages, so that a comparison of the effectiveness of the recommendations may efficiently be demonstrated without requiring any modification to the code or architecture of the web server.

Claims

1. A recommendation appliance for introducing recommendation messages in a web page server system, comprising:

a first port for communicating with at least one end user computer that comprises a web browser for issuing web page requests and displaying requested web pages;

a second port for communicating with at least one web server computer that generates requested web page content in response to a web page request;

at least one processing device; and

a memory having stored thereon executable instructions and data which, when executed by the processing device, cause the recommendation appliance to;

forward a web page request received from a first end user computer at the first port through the second port to a first web server computer identified in the web page request;

receive at the second port a requested web page comprising original web page content from the first web server computer that is generated in response to the web page request from the first end user computer;

generate at least one purchase recommendation based on a recommendation context that is extracted from the web page request or from the requested web page received from the first web server computer;

modify the original web page content to include the purchase recommendation, thereby generating a modified web page; and

return the modified web page through the first port to the first end user computer.

2. The recommendation appliance of claim 1, where the recommendation appliance comprises a reverse proxy server.

3. The recommendation appliance of claim 1, where the recommendation appliance comprises a reverse proxy server where the first port is operatively coupled to a load balancing server and the second port is operatively coupled to a web server.

4. The recommendation appliance of claim 1, where the recommendation appliance comprises a reverse proxy server where the first port is operatively coupled to a first load balancing server and the second port is operatively coupled to a second load balancing server which is operatively coupled to a plurality of web servers.

5. The recommendation appliance of claim 1, where the purchase recommendation is a set of instructions for display with the requested web page by the web browser at the first end user computer.

6. The recommendation appliance of claim 1, where the executable instructions and data which cause the recommendation appliance to modify the original web page content by deleting, inserting or modifying all or part of the original web page content in real time before being returned to the first end user computer.

7. The recommendation appliance of claim 1, where the purchase recommendation comprises static content.

8. The recommendation appliance of claim 1, where the purchase recommendation comprises interactive content.

9. The recommendation appliance of claim 8, further comprising executable instructions and data which cause the recommendation appliance to respond to any response from the first end user computer in response to the interactive content that was not included in the original web page content.

10. The recommendation appliance of claim 1, where the purchase recommendation comprises passive content.

11. The recommendation appliance of claim 1, further comprising executable instructions and data which cause the recommendation appliance to:

separate a plurality of web page requests into at least a first group of web page requests and a second group of web page requests on a predetermined basis, where requested web pages generated by the first group of web page requests are modified to include a first purchase recommendation, and requested web pages generated by the second group of web page requests are modified to include a second purchase recommendation; and

separately record transaction activities of the first group of web page requests and the second group of web page requests.

12. The recommendation appliance of claim 11, where the first purchase recommendation is based on a first recommendation technique and where the second purchase recommendation is based on a second recommendation technique.

13. The recommendation appliance of claim 1, further comprising executable instructions and data which cause the recommendation appliance to:

separate a plurality of web page requests into at least a first group of web page requests and a second group of web page requests on a predetermined basis, where requested web pages generated by the first group of web page requests are modified to include a first purchase recommendation, and requested web pages generated by the second group of web page requests are not modified; and

separately record transaction activities of the first group of web page requests and the second group of web page requests.

14. The recommendation appliance of claim 1, further comprising executable instructions and data which cause the recommendation appliance to:

separate a plurality of web page requests into at least a first group of web page requests and a second group of web page requests on a predetermined basis, where the first group of web page requests is to receive modified web pages and the second group of web page requests is to receive requested web pages that are not modified; and

separately record transaction activities of the first group of web page requests and the second group of web page requests.

15. The recommendation appliance of claim 1, where the first port comprises a wireless connection.

16. A computer-based method of generating recommendations at an integrated recommendation proxy server for potential purchase by a customer, comprising:

receiving a web page request from an client computer;

forwarding the web page request to a web server identified in the web page request;

generating a purchase recommendation based on a recommendation context associated with the web page request while the web page request is forwarded to the web server;

receiving a requested web page comprising original web page content from the web server that is generated in response to the web page request;

modifying the original web page content to include the purchase recommendation, thereby generating a modified web page; and

returning the modified web page to the client computer.

17. The method of generating recommendations of claim 16, where generating a purchase recommendation comprises selecting from among a plurality of association rules to identify a plurality of candidate recommendations having a trigger that matches the recommendation context associated with the web page request.

18. The method of generating recommendations of claim 16, further comprising modifying the web page request to adapt any part of the web page request that corresponds to previously modified functionality or content such that the web server identified in the web page request can interpret the web page request.

19. The method of generating recommendations of claim 16, further comprising modifying the web page request to adapt any part of the web page request that corresponds to a purchase recommendation included in a previously returned modified web page so that the web server identified in the web page request can interpret the web page request.

20. The method of generating recommendations of claim 16, where the recommendation context comprises order history information associated with a user who issued the web page request.

21. A system for generating purchase recommendations for a user, comprising:

a first recommendation appliance comprising

means for receiving a web page request from a user and forwarding the web page request to a web server identified in the web page request;

means for identifying recommendation context information associated with the received web page request;

means for generating a purchase recommendation based on the recommendation context information after the web page request is sent to the web page server;

means for receiving a requested web page generated by the web page server in response to the web page request; and

means for modifying the requested web page to include the purchase recommendation.

22. The system of claim 21, where the means for generating a purchase recommendation comprises:

a database for storing a plurality of association rules;

means for identifying a plurality of candidate recommendation rules from the association rules that match a recommendation context associated with the user;

means for selecting the purchase recommendation from the plurality of candidate recommendation rules.

23. The system of claim 21, further comprising a second recommendation appliance connected in between the first recommendation appliance and the user, where said second recommendation appliance may be configured to modify a modified requested web page received from the first recommendation appliance.