Biased recommender system based on age parameter
The present invention relates to an apparatus, a method and a computer program product for controlling a recommender system, wherein a bias is applied to the output of a recommender in order to favour newly added services and content sources. The bias may decay in time and/or with the number of ratings given to content coming from the new service.
1. A biased recommender apparatus for controlling a recommender system, said apparatus comprising a signal processing device, a computer device, or a chip, and at least one memory including a software routine or program, the at least one memory and the software routine or program configured to with the signal processing device, computer device, or chip, cause the apparatus to:
a) calculate, based on information data of a content item together with data indicating like or dislike of a user, a predicted rating value for a content item received from a content source;
b) determine an age of subscription to said content source so as to derive an age parameter of said content source; and
c) apply a correction value to said predicted rating value based on said age parameter in order to favour content items of new content sources.
2. The apparatus according to claim 1 , wherein said determining an age of subscription determines a subscription time based on metadata, and calculates said age parameter based on said subscription time and an actual time.
3. The apparatus according to claim 2 , wherein said determining an age of subscription determines said subscription time by accessing a look-up table.
4. The apparatus according to claim 1 , wherein said applying a correlation value to said predicted rating value decreases said correction value with increasing age parameter or with increasing number of rating values for received content items of said content source.
5. The apparatus according to claim 4 , wherein said applying a correlation value to said predicted rating value decreases said correction value in a linear or exponential relation to time or said number of rating values.
6. The apparatus according to claim 1 , wherein said calculating a predicted rating value calculates said predicted rating value in a range [0, 1].
7. The apparatus according to claim 1 , wherein said age parameter is a novelty value which indicates how long said content source has been available in said recommender system.
8. The apparatus according to claim 7 , wherein said novelty value is a positive real number less or equal to 1.
9. The apparatus according to claim 1 , wherein said correction value results in a higher rating value for younger subscriptions.
10. A method of controlling a recommender system, said method comprising:
a) calculating, based on information data of a content item together with data indicating like or dislike of a user, a predicted rating value for a content item received from a content source;
b) determining an ace of subscription to said content item so as to derive an age parameter of said content source; and
c) applying a correction value to said predicted rating value based on said age parameter in order to favour content items of new content sources.
11. The method according to claim 10 , further comprising triggering an update of a content pre-selection filter in response to an addition of a new content source.
12. The method according to claim 10 , further comprising triggering in response to an addition of a new content source, a user input operation for creating a new channel with pre-selection filter to the new content source.
13. A computer program product comprising producing signal processing device, or computer device, or chip, and at least one memory including a software routine or program, the at least one memory and the software routine or program configured to with the signal processing device or computer device, cause the computer program product at least to perform the steps of method claim 10 .