| 主权项 |
1. A method of presenting personalized recommendation results to a user at a social networking site with a computing system, comprising:
(a) processing user profile information for the user by the computing system to identify a first set of user content predicted by a recommendation engine to be liked by the user; (b) by the computing system, identifying a second set of users who are designated by the user as friends of the user at the social networking site, and an engagement rating by the user with each user of the second set of users, wherein the engagement rating measures interaction level of the user towards each user of the second set of users in a pre-defined period of time; (c) by the computing system, identifying actions taken by the second set of users in reviewing a second set of user content associated with the second set of users; (d)(1) by the computing system, identifying explicit content ratings given by the second set of users for the second set of user content, and (d)(2) by the computing system, weighing each content item of the second set of user content using both a time recency value for the content item and the engagement rating for a respective friend who rated the content item, wherein the time recency value indicates how recently the respective friend rated the content item; (e) by the computing system, processing the first set of user content, the weighted second set of user content and the explicit content ratings using the recommendation engine to generate a list of top ranked recommended content predicted to be liked by the user; and (f) by the computing system, presenting one or more content items from the list of top ranked recommended content within a graphical user interface for the user at the social networking site during an Internet session. |
