Personalized Movie Hybrid Recommendation Model Based on GRU

the video platform is rich in movie resources, and the huge amount of data has obvious information overload problem. It is a difficult problem for users to quickly find interesting movie resources in the massive data, and the movie recommendation model is an effective way to solve this problem. It is considered to use the deep neural network DNN and factor decomposition machine FM to process the characteristics of users' demands for different movie types, and use GRU (Gate Recurrent Unit) to mine the laws and characteristics of their existence from the text data of movies. In order to make the recommendation result more accurate, a movie recommendation model based on GRU, DNN and FM is proposed by combining the relationship between users and movies. Finally, a large number of experiments are carried out on the Movielens-1m dataset. The experimental results show that compared with LSTM (Long-term and Short-term Memory Network) and CNN (Convolutional Neural Network) recommendation models, the model has better performance and higher recommendation accuracy.