US 7,403,935 B2
Training, inference and user interface for guiding the caching of media content on local stores
Eric J. Horvitz, Kirkland, Wash. (US); Carl M. Kadie, Bellevue, Wash. (US); Stuart Ozer, San Francisco, Calif. (US); and Curtis G. Wong, Bellevue, Wash. (US)
Assigned to Microsoft Corporation, Redmond, Wash. (US)
Filed on May 03, 2005, as Appl. No. 11/121,219.
Application 11/121219 is a continuation of application No. 09/825800, filed on Apr. 04, 2001, granted, now 6,947,935.
Prior Publication US 2005/0193414 A1, Sep. 01, 2005
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 17/30 (2006.01); G06F 13/00 (2006.01)
U.S. Cl. 707—1  [707/100; 707/200; 725/46; 725/91] 17 Claims
OG exemplary drawing
 
1. A system comprising computer executable instructions embodied on a computer readable storage medium for storing media content locally that is received from an information delivery system for viewing at a local system, comprising;
a database system that logs user selections of previously viewed media content at a local system received from an information delivery system;
an inference system, trained by the logged selections of previously viewed media content, that assigns preference values to media content selections in a recommendation list based on the previously viewed media content; and
a local storage system that automatically chooses one or more media content selections based upon the assigned preference values and stores the chosen one or more media content selections, the local storage system assigns a value density to each of the media content selections based on the assigned preference value of the media content selection divided by a storage size of the media content selection, the local storage system receives new media content selections with assigned preference values and dynamically adjusts location of the stored media content based on the new media content selections in the local storage system, wherein media content with the lowest value density is aged out of the local storage system until there is sufficient storage space on the local storage system to store the new media content selections.