E-Commerce Products Personalized Recommendation Based on Deep Learning

Abstract

To solve the problem that the recommendation accuracy of electronic products is low, a personalized recommendation model for e-commerce products based on BERT-BiLSTM was proposed. The two pre-training tasks in the BERT model were used to realize the bidirectional language model. Then, on this basis, the bidirectional neural network BLSTM was introduced to obtain the contextual semantic information of the text, which was output after combining the output of forward and backward hidden layers. Experimental results showed that compared with the benchmark model, BERT-SVM model, BERT-RNN model and BERT-LSTM model, the RMSE value of personalized recommendation model for e-commerce products based on BERT-BiLSTM is the lowest, which is 0.82, which means that the recommendation accuracy of the proposed model is the highest. Therefore, the proposed model is feasible in personalized recommendations for ecommerce products.