Interest Point Recommendation based on Multi Feature Representation and Attention Mechanism

A deep Interest point recommendation model based on multi feature representation and attention mechanism is proposed. On the basis of deep recommendation method, the extraction model of Interest point and user characteristics was constructed. Specifically, the constraint matrix decomposition method was used to obtain the joint representation of geographic location and category features of Interest point, and the semantic features of Interest point were obtained by word embedding model to realize the comprehensive and in-depth multi feature representation of Interest point; At the same time, the generation method of user characteristics was improved, and attention mechanism was used to improve the ability of the model to obtain the key information of user preference; In addition, a deep learning recommendation framework is constructed to model the nonlinear interaction between features to improve the accuracy of recommendation.