AUTOMATIC DETERMINATION OF TYPICAL FOR ARTISTS RELEVANCY OF RECOMMENDATIONS IN SOCIAL NETWORK Russian patent published in 2016 - IPC G06F7/10 H04L12/28 

FIELD: information technology.

SUBSTANCE: invention relates to an operation method of a recommendation system. Idea of the invention is automatic training of the user in a social network aimed at explaining which recommendations of the user contacts also being members of the social network are relevant with respect to the category, in which the user is interested. Instructional algorithm is used to interpret feedback from the user in response to reception of recommendations from his/her contacts. Thus, for each combination of a contact and category there can be determined a "relevancy-taste" index. Certain "relevancy-taste" index is filtered. Only such recommendations are provided to the users associated "relevancy-taste" indices of which satisfy the filtration criterion. So, it becomes possible to considerably reduce the number of irrelevant recommendations presented to the user.

EFFECT: technical result consists in reduction of the number of irrelevant recommendations presented to the user.

14 cl, 2 dwg, 1 tbl