Rules recommendation based on customer feedback

Patent metadata

Field	Value
Publication number	US-9984386-B1
Application number	US-201514709227-A
Country	US
Kind code	B1
Filing date	May 11, 2015
Priority date	May 11, 2015
Publication date	May 29, 2018
Grant date	May 29, 2018

Abstract

Methods and systems for managing or regulating offer listings in an electronic marketplace. Consumer feedback regarding the offer listings may be received and analyzed to generate recommended or candidate rules. The recommended rules may be presented to rule authors for validation. Acceptance and/or rejection of the recommended rules can be used to train or otherwise improve the generation of future recommended rules. The accepted recommended rules can be used to generate final rules. The final rules may be applied to offer listings to trigger a variety of actions including filtering of offer listings, sending notifications to sellers, and updating offer listings user interface.

First claim

What is claimed is:

1. A computer-implemented method, comprising:
providing an offer listings user interface configured to present an offer for an item, wherein the offer includes a plurality of offer attributes including at least a price and a seller, the offer listing user interface including a user feedback interface for receiving feedback from a user;
receiving, via the user feedback interface, user-provided feedback regarding the offer, the user-provided feedback comprising feedback text;
extracting one or more entities from the user-provided feedback based at least in part on textual analysis of the feedback text;
determining one or more entity relations with respect to the one or more entities based at least in part on domain-specific information associated with the one or more entities and context information associated with the feedback;
generating, based at least in part on at least one of the one or more entity relations, a recommended rule comprising a rule condition and a rule action, the rule action comprising instructions to update an attribute of the offer;
providing the recommended rule to the seller; and
upon receiving an indication from the seller that the recommended rule should be applied, applying the recommended rule to the offer, by at least:
determining whether the rule condition of the at least one recommended rule is satisfied by the offer attributes of the offer; and
in response to determining that the rule condition is satisfied, executing the rule action with respect to the offer.

2. The computer-implemented method of claim 1 , wherein the rule condition of the recommended rule is based on at least one offer attribute associated with the offer.

3. The computer-implemented method of claim 2 , wherein the rule condition is satisfied when the price of the offer exceeds a threshold price extracted from the user-provided feedback.

4. The computer-implemented method of claim 2 , wherein the rule condition is satisfied when the seller of the offer is the same as a seller extracted from the user-provided feedback.

5. A computer-implemented method, comprising:
receiving user-provided feedback from an offer listings user interface configured to display a plurality of offers;
extracting, from the user-provided feedback, one or more entities;
determining one or more entity relations between the one or more entities based at least in part on context information associated with the user-provided feedback;
generating a recommended rule based at least in part on the user-provided feedback, the recommended rule comprising a rule condition based at least in part on the one or more entity relations; and
applying the recommended rule to at least one offer of the plurality of offers, comprising:
determining whether the rule condition of the recommended rule is satisfied by the at least one offer listing; and
in response to determining that the rule condition is satisfied, triggering a predetermined rule action associated with the rule condition of the recommended rule, the rule action causing one or more attributes of at least one offer of the plurality of offers to be updated.

6. The computer-implemented method of claim 5 , wherein the user-provided feedback comprises a user identifier, a reference to an offer associated with the user-provided feedback, and feedback text.

7. The computer-implemented method of claim 5 , wherein generating the recommended rule comprises:
extracting a plurality of entities from the user-provided feedback based on at least in part on textual analysis of the feedback text;
extracting one or more entity relations among the plurality of entities based at least in part on domain-specific information associated with the one or more entities and context information associated with the feedback; and
generating the recommended rule based at least in part on the extracted one or more entity relations.

8. The computer-implemented method of claim 7 , wherein the context information includes information about an item, an item attribute, a seller, a buyer, an offer, or an offer attribute associated with the feedback.

9. The computer-implemented method of claim 5 , wherein the predetermined rule action comprises at least one of (1) removing the at least one offer from the plurality of offers; (2) sending a notification to a seller associated with the at least one offer; or (3) updating the offers listings user interface to display an updated offer attribute associated with the at least one offer.

10. The computer-implemented method of claim 9 , wherein the updated offer attribute includes a condition indicator or a seller reputation indicator.

11. The computer-implemented method of claim 5 , further comprising presenting the recommended rule for validation to a reviewer at a rules selection interface configured to enable a reviewer to apply or discard the recommended rule.

12. The computer-implemented method of claim 11 , wherein the reviewer is a human user.

13. A computer system, comprising:
a memory that stores computer-executable instructions; and
a processor configured to access the memory and execute the computer-executable instructions to at least:
receive a user-provided feedback from an offer listings user interface configured to display a plurality of offers, the user-provided feedback associated with context information;
extract, from the user-provided feedback, one or more entities and one or more entity relations between the one or more entities based at least in part on the context information associated with the user-provided feedback;
generate one or more recommended rules based at least in part on the one or more entity relations, each individual recommendation rule of the one or more recommendation rules comprising a rule condition based at least in part on one or more entity relations of the one or more entity relations;
present the one or more recommended rules with a rules selection user interface; and
generate a final rule based at least in part on one or more reviewer instructions received with the rules selection user interface.

14. The computer system of claim 13 , wherein at least one of the one or more recommended rules is generated based at least in part on more than one user-provided feedback.

15. The computer system of claim 13 , wherein the context information includes information about an item, an item attribute, a seller, a buyer, an offer, or an offer attribute associated with the respective user-provided feedback.

16. The computer system of claim 13 , wherein the processor is configured to access the memory and execute the computer-executable instructions to further apply the final rule to the plurality of offers such that a predetermined action is triggered when the rule condition of the final rule is satisfied by at least one of the plurality of offers.

17. The computer system of claim 16 , wherein the predetermined action includes sending a notification to a seller associated with the at least one offer.

18. The computer system of claim 16 , wherein the predetermined action includes updating an offer attribute for the at least one offer.

19. The computer system of claim 16 , wherein the predetermined action includes rendering the at least one offer invisible from the offer listings user interface.

20. The computer system of claim 13 , generating the one or more recommended rules is further based on a previous reviewing instruction received from the rules selection user interface.

Assignees

• Amazon Tech Inc

Inventors

• Bhatia Sandeep

Classifications

• G06Q30/0241 Primary

  Advertisements · CPC title