Appeal 2007-1122 Application 09/966,414 for each subscriber for the available digital items (step 154), determines which of those items, based on their ratings, are likely to interest the subscriber, and generates a new list 44 of recommended items (step 156). (Col. 8, l. 63 to col. 9, l. 1; Fig. 7a.) The collaborative filter 42 transfers the additions to the list and their respective ratings to the scheduling processor 46 (step 158). (Col. 9, ll. 1-3; Fig. 7a.) 8. In the preferred embodiment, the collaborative filter 42 receives a subscriber profile update from the local server. (Col. 9, ll. 4-6; Fig. 7b.) The collaborative filter 42 then updates the subscriber profile 40 (step 164), recomputes the subscriber similarity groups (step 166), predicts subscriber ratings (168), and determines revisions to the subscriber's list 44 of recommended items (step 170). (Col. 9, ll. 7- 11; Fig. 7b.) The filter 42 sends rating changes to the subscriber (step 172) and sends additions to the scheduling processor (step 174). (Col. 9, ll. 11-13; Fig. 7b.) 9. A user may select a recommended item from a list 44 of recommended items displayed on a subscriber interface 58. (Col. 6, ll. 25-31.) After the user uses the requested item, the subscriber interface 58 preferably prompts the subscriber to enter a rating. (Col. 6, ll. 36-38.) 7Page: Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 Next
Last modified: September 9, 2013