Time-centric training, interference and user interface for personalized media program guides

• US 7,644,427 B1
• Filed: 01/31/2005
• Issued: 01/05/2010
• Est. Priority Date: 04/04/2001
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Horvitz, Eric J., Ozer, Stuart, Kadie, Carl M.
• Primary Examiner(s): Saltarelli; Dominic D
• Application Number:
  US11/047,169
• Time in Patent Office: 1,800 Days
• Field of Search: 725/46, 725/47, 725 9- 15
• US Class Current: 725/46
• CPC Class Codes: H04H 60/37 for identifying segments of broadcast information, e.g. scenes or extracting programme ID
• CPC Class Codes: H04H 60/40 for identifying broadcast time
• CPC Class Codes: H04H 60/64 for providing detail information
• CPC Class Codes: H04N 21/252 Processing of multiple end-users' preferences to derive collaborative data
• CPC Class Codes: H04N 21/44222 Analytics of user selections, e.g. selection of programs or purchase activity monitoring of user selections in data processing systems G06F11/34; arrangements for monitoring the user's behaviour or opinions in broadcast systems H04H60/33
• 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 21/4662 characterized by learning algorithms
• CPC Class Codes: H04N 21/4668 for recommending content, e.g. movies
• CPC Class Codes: H04N 21/4756 for rating content, e.g. scoring a recommended movie
• CPC Class Codes: H04N 7/17318 Direct or substantially direct transmission and handling of requests

Abstract

The present invention is related to a system and method of considering time segments or intervals in a collaborative filtering model. The present invention extends collaborative filtering approaches by integrating considerations of temporality into the training and/or vote input associated with the usage of collaborative filtering models. The present invention also applies filtering to the output with temporal models, so as to view a most appropriate subset of recommended content, centering on content that may be available at a target time. The present invention applies time to a collaborative filtering model by allowing weight to be associated with selections within a current time segment, selections historically watched within the current time segment by the user and selections historically watched within the current time segment by a large group of users.

Claims

1. A method for ranking recommendations of multimedia program items in a selectable items list received from an information delivery system, wherein the method is implemented by a processor executing computer-executable instructions stored in computer-readable media, the method comprising:

establishing a threshold dwell time period, wherein when a viewer views a particular multimedia program item for longer than the threshold dwell time period, the particular multimedia program item is assigned a positive vote, wherein when the viewer views the particular multimedia program item for less than the threshold dwell time period, the particular multimedia program item is assigned a negative vote;

logging, by the processor, titles of multimedia program items viewed by a plurality of viewers and temporal history corresponding to the viewed multimedia program items in a database, wherein the temporal history is at least one time segment of a day or a week, wherein each multimedia program item that is viewed for more than the threshold dwell time period is logged with a positive vote, wherein each multimedia program item that is viewed for less than the threshold dwell time period is logged with a negative vote;

training a collaborative filtering system using the logged temporal history of the viewed multimedia program items, wherein the training takes into account both positive and negative votes logged based on the threshold dwell time period for multimedia program items during particular time intervals; and

providing a changeable and selectable recommendation list for display in an electronic program guide in a television system, wherein the recommendation list is ordered by an estimated degree of preference for a viewer based on the trained collaborative filtering system and according to a target time period for viewing the multimedia program items.

2. The method of claim 1, further comprising providing a list of previously viewed multimedia program items to be displayed to the viewer, wherein the list of previously viewed multimedia program items includes an indication of whether each previously viewed multimedia program item listed received a positive vote or a negative vote.

3. The method of claim 2, further comprising

receiving a selection from the viewer to change a positive vote on the list of previously viewed multimedia program items to a negative vote or to change a negative vote to a positive vote for one or more of the previously viewed multimedia program items listed;

and

updating the recommendation list in response to the selection received to change the vote for one or more of the previously viewed multimedia program items.

4. The method of claim 1, the step of logging titles of multimedia program items viewed by a plurality of viewers and temporal history corresponding to the viewed multimedia program items in a database comprising time stamping the titles of the viewed items according to event type and the collaborative filtering system comprising a single collaborative filtering model adapted to be trained based on time intervals during which the multimedia program items have been viewed.

5. The method of claim 1, the step of training the collaborative filtering system comprising utilizing a plurality of separate collaborative filtering models, each collaborative filtering model being trained with the multimedia program items from a particular time interval of temporal history that has been viewed.

6. The method of claim 1, the step of logging titles of multimedia program items viewed by a plurality of viewers and temporal history corresponding to the viewed multimedia program items in a database comprising time stamping the titles of the viewed multimedia program items according to event occurrence.

7. The method of claim 1, the collaborative filtering system comprising a single collaborative filtering model adapted to be trained based on time intervals during which the multimedia program items were viewed.

8. The method of claim 1, further comprising utilizing other attributes of one of the plurality of viewers of the viewed multimedia program items in the step of providing a selectable recommendation list ordered by estimated degree of preference that a viewer has for each multimedia program item.

9. The method of claim 1, the step of providing a selectable recommendation list of recommended multimedia program items ordered by estimated degree of preference that a viewer has for each multimedia program item comprising providing a globally ranked selection list to a cluster of systems based on the temporal viewing history of the other viewing systems of the cluster.

10. The method of claim 1, further comprising filtering a reviewed multimedia program items list, inputting the filtered reviewed multimedia program items list into the trained collaborative filtering system and generating a new selectable recommendation list of titles according to the preferences provided by the filtered reviewed multimedia program items list.

11. The method of claim 10, the step of filtering comprising applying at least one of a live show selection filter, a time period filter, a popularity filter, a pattern search engine and an adding multimedia program items of interest and update component to the reviewed multimedia program items list.

12. A filtering system for ranking recommendations of multimedia program items in a selectable list of multimedia program items received from an information delivery system, comprising:

a processor coupled to computer readable media, the computer readable media storing instructions adapted to be executed by the processor for;

logging titles of multimedia program items viewed by two or more viewers of a shared system and temporal history corresponding to the viewed multimedia program items in a database, wherein the temporal history is one or more time segments, wherein each multimedia program item that is viewed for more than a threshold dwell time period is logged with a positive vote, wherein each multimedia program item that is viewed for less than the threshold dwell time period is logged with a negative vote;

retrieving the logged titles and the temporal history corresponding to the viewed multimedia program items and vote information from the database;

employing collaborative filtering techniques on the temporal history of the viewed multimedia program items and a selection list of available titles to order an editable recommendation list by estimated degree of preference that a present viewer has for each item, wherein the ordering takes into account both positive and negative votes logged based on the threshold dwell time period for multimedia program items during particular time intervals; and

providing the editable recommendation list to the present viewer of a viewing system.

13. The filtering system of claim 12, wherein the processor provides a list of previously viewed multimedia program items to be displayed to the viewers, wherein the list of previously viewed multimedia program items includes an indication of whether each previously viewed multimedia program item listed received a positive vote or a negative vote collaborative filtering techniques further comprise utilizing other attributes of the user of the viewed information.

14. The filtering system of claim 12, wherein the processor receives a selection from one or more of the viewers to change a positive vote on the list of previously viewed multimedia program items to a negative vote or to change a negative vote to a positive vote for one or more of the previously viewed multimedia program items listed, and updates the recommendation list in response to the selection received to change the vote for one or more of the previously viewed multimedia program items collaborative filtering techniques further comprising for utilizing attributes of other viewing systems.

15. The filtering system of claim 12, wherein the collaborative filtering techniques further comprise providing a globally ranked selection list to a cluster of viewing systems in communication with the filtering system based on the temporal viewing history of other viewing systems of the cluster.

16. The filtering system of claim 12, further comprising filtering a reviewed item list, the filtered reviewed items list being a preference input into the filtering system for employing collaborative filtering techniques.

17. The filtering system of claim 12, wherein a viewing system comprising a television receives the ranked selectable list in an electronic program guide.

18. A method for providing a personalized recommendation list of titles of multimedia program items, wherein the method is implemented by a processor executing computer-executable instructions stored in computer-readable media, the method comprising:

establishing a threshold dwell time period, wherein when a viewer views a particular multimedia program item for longer than the threshold dwell time period, the particular multimedia program item is assigned a positive vote, wherein when the viewer views the particular multimedia program item for less than the threshold dwell time period, the particular multimedia program item is assigned a negative vote;

logging titles of multimedia program items viewed by a viewer of a viewing system and temporal history corresponding to the viewed information in a database, wherein the temporal history is a time interval during which the information is viewed, wherein each multimedia program item that is viewed for more than a threshold dwell time period is logged with a positive vote, wherein each multimedia program item that is viewed for less than the threshold dwell time period is logged with a negative vote;

retrieving the logged titles and the temporal history corresponding to the viewed multimedia program items and vote information from the database;

employing collaborative filtering techniques on the temporal history of the viewed information and a selection list of available titles to order an editable recommendation set by estimated degree of preference that a current viewer has for each item, wherein the recommendation list is selectable by time segment, wherein the ordering takes into account both positive and negative votes logged based on the threshold dwell time period for multimedia program items during particular time intervals;

providing the editable recommendation set to the current viewer of the viewing system;

providing a list of previously viewed multimedia program items to be displayed to the current viewer, wherein the list of previously viewed multimedia program items includes an indication of whether each previously viewed multimedia program item listed received a positive vote or a negative vote;

receiving a selection from the current viewer to change a positive vote on the list of previously viewed multimedia program items to a negative vote or to change a negative vote to a positive vote for one or more of the previously viewed multimedia program items listed; and

updating the recommendation list in response to the selection received to change the vote for one or more of the previously viewed multimedia program items.