摘要
大数据时代,必然涌现出各种各样的海量数据,而推荐系统是帮助人们选择数据的有效手段之一。目前,以协同过滤算法为代表的传统推荐算法已经无法满足人们的个性化选择的需求。本文利用深度神经网络构建基于深度学习的推荐模型,抽取用户和电影的特征,并且设计一个多层神经网络将用户和电影特征进行深度交互,从而挖掘用户和电影的深层交互关系,得出用户的偏好。通过相关Spark、Flink、Tensorflow等技术实现对深度学习电影推荐系统的构建和部署。研发出了个性化电影推荐系统。
In the era of big data, various kinds of massive data could inevitably emerge. Recommendation system is one of the effective means to help people choose data. Nowadays, the traditional recommendation algorithm represented by collaborative filtering algorithm can not meet the needs of people′s personalized selection. This paper uses deep neural network to build a recommendation model based on deep learning for extracting the features of users and movies, and designs a multi-layer neural network to conduct deep interaction between the features of users and movies, in order to mine the deep interaction relationship between users and movies, and obtaining the preferences of users. The construction and deployment of the in-depth learning movie recommendation system are realized through relevant Spark, Flink, TensorFlow and other technologies. A personalized movie recommendation system is developed.
作者
梁肇敏
梁婷婷
LIANG Zhaomin;LIANG Tingting(College of Artificial Intelligence,Nanning University,Nanning 530200,China)
出处
《智能计算机与应用》
2022年第10期157-162,共6页
Intelligent Computer and Applications
基金
广西高校中青年教师科研基础能力提升项目(2021KY1800)
南宁学院校级科研项目(2020XJ10)。
关键词
深度神经网络
推荐系统
深度学习
deep neural network
recommendation system
deep learning
