FIELD: information technology.
SUBSTANCE: method is performed by receiving a request for content reccomendation from the electronic device, the content recommendation includes recommended content element, executing the module of the first machine learning algorithm to determine a subset of recommended content sources from a variety of possible content sources, determining a subset of the recommended content sources, including receiving indications of previous interactions of users with at least one of: the system reccomendations and at least some of the plurality of network resources, based on previous user interactions, determining the first subset of content sources associated with the user, based on the trained formula of machine learning algorithm of other user interactions with some other of the plurality of network resources, and analysis of a subset of recommended content sources for the selection of potentially recommended content items.
EFFECT: increasing the effectiveness of recommendations for content available on network resources.
22 cl, 5 dwg
