US 12,277,161 B2
Content recommendation system and method
Peter Docherty, Glasgow (GB); Christopher McGuire, Glasgow (GB); Xiaowei Zhang, Glasgow (GB); Richard Dowling, Glasgow (GB); and Craig Palmer, Redondo Beach, CA (US)
Assigned to ThinkAnalytics Ltd., Lanarkshire (GB)
Filed by ThinkAnalytics Ltd., Lanarkshire (GB)
Filed on Apr. 4, 2023, as Appl. No. 18/130,899.
Prior Publication US 2024/0338404 A1, Oct. 10, 2024
Int. Cl. G06F 7/00 (2006.01); G06F 16/00 (2019.01); G06F 16/435 (2019.01)
CPC G06F 16/435 (2019.01) 18 Claims
OG exemplary drawing
 
1. A computer-implemented method of providing one or more content item recommendations for a user of a content distribution system, comprising:
determining if a user search using a title of content finds the title of content in one or more content sources available to the user, the one or more content sources having content available for the user to watch or otherwise consume;
identifying, in response to the title of content not being found in the one or more content sources in that the content searched for by the user is not available for the user to watch or otherwise consume, the content in a content library by determining if the title of content is in the content library, wherein the content library is a different database from the one or more content sources available to the user in that the content library contains content that is not available to the user to watch or otherwise consume and wherein the content library contains metadata concerning items of content in the content library, the metadata representing at least some properties of the items of content;
using, in response to the title of content being determined to be in the content library, the metadata of the identified content in the content library and metadata concerning content available from the one or more content sources available to the user to watch or otherwise consume to generate at least one content item recommendation for the user, wherein the at least one content item recommendation is for recommended content that is in the one or more content sources available to the user and has not been searched for by the user, and that is related to the identified content in that it is similar to the content that has been searched for by the user;
providing the at least one content item recommendation to the user; and
producing a reduced content library by matching candidate titles of content to titles of content in the extended content library, and adding the metadata of the matched content from the extended content library to the reduced content library.