Method for using only feedback from shows that improve the performance of the recommender system
View Patent ↗In customizing a user profile employed by a recommendation system, users are prompted for feedback regarding content that is the subject of the recommendation system. Only feedback that does not degrade performance of the recommendation system, as measured by the error rate, is accepted and utilized to modify the user profile. Feedback that would degrade performance is discarded without being employed to alter the user profile. In this manner, error is continually driven toward a minimum by system changes based on feedback.
1. A system for customizing a user profile for a recommendation system comprising:
a hardware controller selectively receiving user feedback relating to content forming the subject for the recommendation system and checking an error rate for a user profile with modifications based upon the user feedback with an error rate for the user profile without the modifications, wherein
the controller only permanently incorporates the modifications in the user profile if the error rate is not degraded by the modifications.
2. The system according to claim 1 , wherein the controller discards the modifications based upon the user feedback if the error rate is degraded by the modifications.
3. The system according to claim 1 , wherein the controller tests the error rate utilizing the user profile with one or more test cases with and without the modifications.
4. The system according to claim 3 , wherein the controller tests the error rate utilizing a correlation between past recommendations and user selections.
5. The system according to claim 1 , wherein the controller accepts the modifications to the user profile based upon the user feedback only if the error rate is improved by the modifications.
6. A system for customizing a user profile for a recommendation system comprising:
a video receiver;
a video display device;
a remote control operative to selectively control operation of the video receiver, the video display device, or both; and
a controller within one or more of the video receiver, the video display device, and the remote control, the controller selectively receiving user feedback relating to content forming the subject for the recommendation system and checking an error rate for a user profile with modifications based upon the user feedback with an error rate for the user profile without the modifications,
wherein the controller only permanently incorporates the modifications in the user profile if the error rate is not degraded by the modifications.
7. The system according to claim 6 , wherein the controller discards the modifications based upon the user feedback if the error rate is degraded by the modifications.
8. The system according to claim 6 , wherein the controller tests the error rate utilizing the user profile with one or more test cases with and without the modifications.
9. The system according to claim 8 , wherein the controller tests the error rate utilizing a correlation between past recommendations and user selections.
10. The system according to claim 6 , wherein the controller accepts the modifications to the user profile based upon the user feedback only if the error rate is improved by the modifications.
11. A method of customizing a user profile for a recommendation system comprising:
selectively receiving user feedback relating to content forming the subject for the recommendation system;
checking, by a hardware controller, an error rate for a user profile with modifications based upon the user feedback with an error rate for the user profile without the modifications; and
permanently incorporating the modifications in the user profile only if the error rate is not degraded by the modifications.
12. The method according to claim 11 , wherein the modifications based upon the user feedback are discarded if the error rate is degraded by the modifications.
13. The method according to claim 11 , wherein the error rate is tested utilizing the user profile with one or more test cases with and without the modifications.
14. The method according to claim 13 , wherein the error rate is tested utilizing a correlation between past recommendations and user selections.
15. The method according to claim 11 , wherein the modifications to the user profile based upon the user feedback are accepted only if the error rate is improved by the modifications.