Research On Personalized Hybrid Recommendation Model Based On Deep Learning

Aiming at the problem of data sparsity in traditional collaborative filtering algorithms, a personalized hybrid recommendation model based on deep learning is proposed. The model improves the local similarity of convolutional neural network. The model uses convolutional neural network and adds an adjustment layer to make the user's interest preference locally characterized on the basis of iterative adjustment of user and item scoring matrix, Then the missing score is predicted. Experiments show that the method proposed in this paper effectively alleviates the data sparsity in the process of personalized recommendation.