Research And Application Of Personalized Recommendation Algorithm Based On Deep Learnin

At the time of the horn of industrialization 4.0,the Internet ushered in the era of information bangs,and the media industry,represented by movies,music and short videos,has also entered the era of intelligence,with users increasingly expecting Internet platforms to provide them with personalised recommendation services.Traditional collaborative filtering recommendation algorithms are not ideal because they cannot adequately collaborate the complex features between users and items,while deep learning has the ability of non-linear learning,which can better explore the complex features between users and items,thus improving the recommendation accuracy.The data sparsity problem and the cold start problem are hot research topics in the field of recommendation algorithms.This paper improves the recommendation accuracy of recommendation systems from two aspects: data sparsity and the cold start problem,and the main research contents are as follows.(1)To address the problem of data sparsity,a deep matrix decomposition recommendation algorithm and a deep hybrid recommendation algorithm are proposed in this paper,and the two methods are validated using two different publicly available datasets.The deep matrix decomposition recommendation algorithm uses the matrix decomposition method in the traditional collaborative filtering idea from the perspective of quantitative information,combined with deep learning neural networks to reduce the dimensionality of the data,which effectively reduces the amount of operations and the impact of data sparsity on the recommendation results.The deep hybrid recommendation algorithm combines convolutional neural networks and recurrent neural networks to build a model based on text filtering recommendations from the perspective of text information.The experimental results show that the deep matrix decomposition recommendation algorithm and the deep hybrid recommendation algorithm can effectively reduce the impact of data sparsity on the recommendation accuracy and improve the accuracy of the recommendation results.(2)For the cold start problem,this paper uses the above proposed deep hybrid algorithm to build a personalized recommendation system for online movies using Django as the framework and Python language.The front-end of the system is designed to implement six functional modules,including user registration and login,popular recommendation,category recommendation,movie rating and review,personalized recommendation and viewing history rating;the back-end is designed to implement user management and database management.The experimental results show that the personalized recommendation system of online movies based on deep hybrid recommendation algorithm can effectively reduce the impact of information overload on users,and has certain engineering applicability and research value.