Training, inference and user interface for guiding the caching of media content on local stores

• US 20050210520A1
• Filed: 05/09/2005
• Published: 09/22/2005
• Est. Priority Date: 04/04/2001
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Wong, Curtis G., Horvitz, Eric J., Ozer, Stuart, Kadie, Carl M.
• Granted Patent: US 7,440,950 B2
• Time in Patent Office: Days
• US Class Current: 725/89
• CPC Class Codes: H04N 21/4147 PVR [Personal Video Recorder] H04N5/76 takes precedence ; arrangements for broadcast with accumulation-type receivers H04H20/40
• CPC Class Codes: H04N 21/4331 Caching operations, e.g. of an advertisement for later insertion during playback
• CPC Class Codes: H04N 21/4335 Housekeeping operations, e.g. prioritizing content for deletion because of storage space restrictions storage management, e.g. defragmentation G06F3/0604; unloading stored programs G06F9/445; storage management in file systems G06F16/10; buffering arrangements in a network node or in an end terminal in packet networks H04L49/90
• CPC Class Codes: H04N 21/44008 involving operations for analysing video streams, e.g. detecting features or characteristics in the video stream arrangements characterised by components specially adapted for monitoring, identification or recognition of video in broadcast systems H04H60/59
• CPC Class Codes: H04N 21/4532 involving end-user characteristics, e.g. viewer profile, preferences monitoring of user activities for profile generation for accessing a video database G06F16/739; user profiles in network data switching protocols H04L67/306; processing of user preferences or user profiles in wireless networks H04W8/18
• CPC Class Codes: H04N 21/466 Learning process for intelligent management, e.g. learning user preferences for recommending movies services using the results of monitoring in broadcast systems H04H60/61
• CPC Class Codes: H04N 21/4667 Processing of monitored end-user data, e.g. trend analysis based on the log file of viewer selections
• CPC Class Codes: H04N 21/47 End-user applications
• CPC Class Codes: H04N 21/4755 for defining user preferences, e.g. favourite actors or genre retrieval personalisation and generation of user profiles for the retrieval of video data G06F16/739; user profiles in network data switching protocols H04L67/306
• CPC Class Codes: H04N 21/482 End-user interface for program selection broadcast systems using EPGs H04H60/72
• CPC Class Codes: H04N 5/76 Television signal recording
• CPC Class Codes: H04N 5/765 Interface circuits between an apparatus for recording and another apparatus
• CPC Class Codes: H04N 5/775 between a recording apparatus and a television receiver
• CPC Class Codes: H04N 7/163 by receiver means only
• CPC Class Codes: Y10S 707/99931 Database or file accessing
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing
• CPC Class Codes: Y10S 707/99937 Sorting
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing

Abstract

The present invention is related to a system and method of caching data employing probabilistic predictive techniques. The system and method has particular application to multimedia systems for providing local storage of a subset of available viewing selections by assigning a value to a selection and retaining selections in the cache depending on the value and size of the selection. The value assigned to an item can represent the time-dependent likelihood that a user will review an item at some time in the future. An initial value of an item can be based on the user'"'"'s viewing habits, the user'"'"'s viewing habit over particular time segment (e.g., early morning, late morning, early afternoon, late afternoon, primetime, late night) and/or viewing habits of a group of user'"'"'s during a particular time segment. A value assigned to a selection dynamically changes according to a set of cache retention policies, where the value can be time-dependent functions that decay based on the class of the item, as determined by inference about the class or via a label associated with the item. A selections value may be reduced as the selection ages because a user is less likely to view the selection over time. Additionally, a value of a selection may change based on changes on a user'"'"'s viewing habits, changes in time segments or a user'"'"'s modification of the cache retention policies.

Claims

1. A multimedia system that stores information received locally from a program delivery system for viewing at a local system, comprising;

a cache loading system that receives a list of selections from a program delivery system and stores program selection information corresponding to the list of selections in a local memory system; and

a utility system that monitors program selection information in the local memory system and communicates value information to the cache loading system for removing information residing in the local memory system in exchange for information having a higher value received by the program delivery system.

2. The system of claim 1, the local memory system comprising a first portion to define user defined selections to be stored, a second portion to define live show selection to be stored and a third portion to define general selections to be stored.

3. The system of claim 1, further comprising a local storage system that receives new selections with assigned values, the utility system dynamically adjusts the location of the stored information and new information based on the new selections in the local storage system, wherein the information is aged out of the local storage system based on an expected value density of each selection, the expected value density determinable based on the assigned value of the selection divided by the size of the selection.

4. The system of claim 1, the utility system dynamically adjusts the assigned values based on a cache retention policy.

5. The system of claim 1, selections are grouped by class types with different cache retention policies for aging the assigned values of selections within different class types, the cache retention policies of each class type is initially defined based on a default time-dependent decay function, the time-dependent decay function being continuously refined based on the likelihood that a user will view a program based on the amount of time that has passed since the program has been recorded and not yet viewed.

6. The system of claim 1, the utility system comprising a time dependent utility model that dynamically adjusts the assigned values based on the time that the selection resides in the local memory system.

7. The system of claim 1, further comprising an inference system that receives a reviewed program list from the program delivery system, provides a recommendation list and dynamically assigns values to selections in the recommendation list based on previously viewed programs, the cache loading system stores the program selection information corresponding to the assigned values of the selections in a local storage system.

8. The system of claim 7, the inference system employing collaborative filtering techniques on a temporal history of previously viewed programs and the selectable program list to dynamically assign values to selections in a recommended program list from a higher probability to a lower probability that a user of the system would prefer to view information corresponding to an available selection, the viewed programs being time stamped by event type and the inference system being based on a single collaborative filtering model that is trained according to time intervals that the information has been viewed.

9. The system of claim 8, the inference system being based on a plurality of separate collaborative filtering models, each collaborative filtering model being trained with the information from a particular time interval of temporal history that has been viewed.

10. The system of claim 8, the viewed programs being time stamped by event occurrence and the inference system being based on a single collaborative filtering model that is trained according to time intervals that the information has been viewed.

11. The system of claim 8, the inference system receives further attributes of at least one user of the system and utilizes these attributes in assigning values to selections in the selectable program list.

12. The system of claim 1, the cache loading system records live shows at any given time period for N number of shows employing N number of tuners, the shows are selected based on the temporal history of previously viewed programs within a time interval covering the any given time period.

13. The system of claim 1, the information system residing on a remote server communicatively coupled to at least one set top box, wherein a recommendation list and information corresponding to programs in the recommendation list are generated by the server and transmitted to the set top box.

14. A multimedia system that locally stores information received from a program delivery system for viewing at a local system, comprising;

a local system having a memory loading system that determines a portion of a multimedia program, downloadable from a remote source to the local system, to store locally based on a local viewing rate and a remote transmission rate of the multimedia program;

a storage system that stores a portion of the multimedia program necessary for uninterrupted viewing of the multimedia program, the storage information stores portions of a plurality of downloadable selections based on a probability that a system user would like to view a selection;

a viewing system that retrieves the multimedia program from the storage system for viewing, wherein a remaining portion of the multimedia program is downloaded to the storage system while the multimedia program is being viewed until the entire multimedia program has been downloaded; and

an inference system that determines the portions of downloadable selections to store in the storage system based on the temporal history of previously viewed programs of the system.

15. The system of claim 14, the inference system employing collaborative filtering techniques on the temporal history and a selectable downloadable program list to dynamically assign values to selections in the selectable program list from a higher probability to a lower probability that a user of the system would prefer to view information corresponding to available selections.

16. The system of claim 15, further comprising a utility system that monitors program selection information in the storage system and communicates aging information to the memory loading system for aging out older program information residing in the storage system in exchange for newer information.

17. A multimedia system that stores television content locally that is received from a program delivery system for viewing at a local system, comprising;

a cache loading system that receives a list of live show selections from a program delivery system and stores the live show selections corresponding to the list of selections in a local memory system; and

a utility system that monitors program selection information in the local memory system and communicates value information to the cache loading system wherein high values are assigned to live show selections currently in progress, which are quickly decayed after the show is no longer live wherein live shows residing in the local memory having lower values are aged out in exchange for live shows having a higher value received by the program delivery system.

18. The system of claim 17, the cache loading system records live shows at any given time period for N number of shows employing N number of tuners, the shows are selected based on the temporal history of previously viewed programs within a time interval.

19. The system of claim 17, the cache loading system records other shows that are not live with the live shows, the live shows are provided with higher initial values than the other shows and aged out quicker when no longer live than the other shows.

20. The system of claim 17, the live show selections beginning at starting standard times, starting at one of the hour and half hour, are provided with higher initial values than the other shows and aged out quicker when no longer live than the other shows, the higher initial values are determined based on the likelihood that the show will be viewed by a user of the system.