| CPC G06F 16/9535 (2019.01) [G06F 16/735 (2019.01); G06F 16/9035 (2019.01)] | 14 Claims |
|
1. A recommended content selection method performed by a computer device, the method comprising:
obtaining an input structure graph, the input structure graph comprising a plurality of user nodes and a plurality of content nodes;
generating, according to the input structure graph, fusion vectors corresponding to nodes in the input structure graph using an interactive prediction model, wherein a fusion vector corresponding to a target node among the nodes integrates feature information of the target node and feature information of a node associated with the target node;
determining, for a target user node among the plurality of user nodes, interactive prediction values between the target user node and the plurality of content nodes based on a fusion vector corresponding to the target user node and fusion vectors corresponding to the plurality of content nodes;
ranking the plurality of content nodes according to a descending order of the interactive prediction values associated with the plurality of content nodes;
selecting, according to the descending order of the interactive prediction values, a top first quantity of candidate content nodes from the plurality of content nodes to obtain a candidate content node set corresponding to the target user node;
selecting a second quantity of candidate content nodes from the candidate content node set as a recommended content node set, the second quantity being less than the first quantity, wherein the recommended content node set has a diversity index higher than any other second quantity of candidate content nodes from the candidate content node set, and the diversity index of the recommended content node set is defined as a determinant of a positive semi-definite matrix composed of the recommended content node set for representing a diversity degree of the recommended content node set;
determining a recommended content based on the recommended content node corresponding to the target user node and correlation between the recommended content and the target user node; and
returning the recommended content to a terminal associated with the target user node, wherein a target application at the terminal is configured to display the recommended content to a target user.
|