Efficient recommendation method based on user session and social network

The invention provides a recommendation method based on a user session and a social network, and the method comprises the following steps: constructing a knowledge graph embedded network: constructing the knowledge graph embedded network which is used for obtaining the user session and is related to the social network by using the social network information and historical behavior information of a user; training and learning a heterogeneous knowledge graph network: establishing the heterogeneous knowledge graph network by utilizing the knowledge graph embedded network, and learning hidden information in users and sessions in the heterogeneous knowledge graph network; embedding personalized session features: aiming at users and session information in the heterogeneous knowledge graph network, utilizing a calculation module to obtain characterization session embedded information aiming at each user, and obtaining personal preference information of the users and context feature vectors in sessions; performing we