System and method for generating effective recommendations

• US 7,720,720 B1
• Filed: 08/05/2004
• Issued: 05/18/2010
• Est. Priority Date: 08/05/2004
• Status: Active Grant

• Current Assignee: Versata Development Group, Inc. (ESW Capital, LLC)
• Original Assignee: Versata Development Group, Inc. (ESW Capital, LLC)
• Inventors: Darr, Timothy P., Packebush, Sherrill J., Devine, Alex, Wacksman, Jeremy S., Grant, Michael K., Sharma, Nirad, Dillon, Thomas H., Thier, Neil C., Hyams, Chris, Singh, James
• Primary Examiner(s): Airapetian; Mila
• Application Number:
  US10/912,734
• Time in Patent Office: 2,112 Days
• Field of Search: 705/26, 705/27
• US Class Current: 705/26.7
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0631 Item recommendations

Abstract

A data processing system generates recommendations for on-line shopping by scoring recommendations matching the customer'"'"'s cart contents product page, configuration page, email and/or history using predetermined scoring criteria, such as item availability, item price, relevance of the item to the customer, item clarity, item exposure, total purchase amount in the cart, margin, recommendation confidence and lift, shopper profile matches, seasonality, customer history, time dependency, product marketing emphasis, substitutability, item popularity and/or product attributes associated with the recommended item. The scored recommendations may be ranked and/or filtered to prioritize the recommendations so that the best recommendation is selected for presentation to the customer. In addition, one or more scoring criteria or product attributes for the recommended item may be used to select a specific selling point/message text for inclusion with the selected recommendation.

Claims

1. A computer-based method of generating recommendations for potential purchase by a customer, comprising:

generating association rules from a transaction history data set;

receiving a recommendation context from a customer;

using the recommendation context at a computer system to identify a plurality of candidate recommendation rules from the association rules that match the recommendation context, where each candidate recommendation rule recommends at least one recommended item;

determining a score for each candidate recommendation rule using a margin value factor for the recommended item, a confidence value factor for the candidate recommendation rule and a predetermined scoring criteria factor;

ranking the plurality of candidate recommendation rules using the score for each candidate recommendation rule to identify at least a highest ranking candidate recommendation rule; and

issuing at least the highest ranking candidate recommendation rule as a recommendation.

2. The method of claim 1, further comprising generating a selling message for at least the highest ranking candidate recommendation rule by selecting from a plurality of selling point messages using an attribute from a customer shopping cart, from a customer history or from an item being recommended by the highest ranking candidate recommendation rule.

3. The method of claim 1, comprising generating a recommendation message to include in a recommendation by using a profile estimate calculated from the recommendation context to select a recommendation message from a table of selling point messages.

4. The method of claim 1, comprising generating a recommendation message by using at least one attribute from each candidate recommendation rule to select a selling message for that candidate recommendation rule from a table of selling point messages.

5. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises adjusting the predetermined scoring criteria factor if a profile estimate calculated from the recommendation context matches a target profile contained within the candidate recommendation rule.

6. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises multiplying the margin value factor, the confidence value factor and the predetermined scoring criteria factor to form the score.

7. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises adjusting the predetermined scoring criteria factor if the recommended item has not previously been purchased by the customer.

8. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises increasing the predetermined scoring criteria factor if the customer has previously researched the recommended item.

9. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises decreasing the predetermined scoring criteria factor if the customer has previously researched the recommended item.

10. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises adjusting the predetermined scoring criteria factor depending on whether a seasonal requirement for the recommended item is satisfied.

11. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises increasing the predetermined scoring criteria factor if there is a marketing emphasis for the recommended item.

12. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises increasing the predetermined scoring criteria factor if the recommended item is a popular item.

13. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises increasing the predetermined scoring criteria factor if the recommended item has a margin exceeding a predetermined margin threshold value.

14. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises increasing the predetermined scoring criteria factor if a combined margin value for the recommended item and any items already purchased by the customer exceeds a predetermined margin threshold value.

15. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises decreasing the predetermined scoring criteria factor if the recommended item has a price exceeding a predetermined price threshold value.

16. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises decreasing the predetermined scoring criteria factor if the recommended item has a price that exceeds a total price for previously purchased items by a predetermined factor.

17. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises decreasing the predetermined scoring criteria factor if a combined price value for the recommended item and any items already purchased by the customer exceeds a predetermined price threshold value.

18. The method of claim 1, further comprising issuing a second candidate recommendation rule as a recommendation, where the second candidate recommendation rule is not substitutable with the highest ranking candidate recommendation rule.

19. The method of claim 1, where the step of determining a score for each candidate recommendation rule comprises multiplying a confidence metric for the candidate recommendation rule by a margin value for the item recommended by the candidate recommendation rule.

20. The method of claim 1, wherein the predetermined scoring criteria is a profile estimate or a message strength value for a selling message associated with the candidate recommendation rule.

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

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 customer, where each candidate recommendation rule recommends at least one recommended item;

means for scoring each candidate recommendation rule by calculating an expected margin value for the recommended item;

means for ranking the expected margin values for the candidate recommendation rules to identify a purchase recommendation for the customer; and

means for generating a selling message for the purchase recommendation by using a user profile to select from a plurality of selling messages based on an attribute of the recommended item associated with the recommended item recommended by the purchase recommendation.

22. An article of manufacture having at least one recordable medium having stored thereon executable instructions and data which, when executed by at least one processing device, cause the at least one processing device to:

use a recommendation context associated with a customer to identify a set of candidate association rules that match a recommendation context associated with a customer, where each candidate association rule recommends at least one recommended item;

calculate a score value for each candidate association rule using an expected margin value for each recommended item and one or more predetermined scoring criteria;

rank the set of candidate association rules using the score value for each candidate association rule; and

issue at least the candidate association rule having the highest score value as a purchase recommendation to a customer.