融合图神经网络的深度学习电影推荐系统设计与实现  

Design and Implementation of Deep Learning Movie Recommendation System Based on Graph Neural Network

作　　者：冯萍[1] 钱阳 李国梁 刘笑涵 FENG Ping;QIAN Yang;LI Guo-liang;LIU Xiao-han(College of Computer Science and Technology,Changchun University,Changchun University,Changchun 130022,China;College of Network Security,Changchun University,Changchun 130022,China)

机构地区：[1]长春大学计算机科学与技术学院,长春130022 [2]长春大学网络安全学院,长春130022

出　　处：《白城师范学院学报》2021年第5期49-56,共8页Journal of Baicheng Normal University

基　　金：吉林省教育厅“十四五”科学研究规划项目“面向老年人慢性病的医疗知识图谱构建与应用研究”(2021LY505L16).

摘　　要：在大数据时代背景下,传统的协同过滤推荐算法已经很难满足人们的个性化推荐需求,针对该问题,文章提出一种融合图神经网络的深度学习推荐模型(GCN-MLP).该模型应用图神经网络推理用户和电影间的交互关系,设计多层感知机挖掘多特征域间复杂的组合关系,并将两者结合得到相应的用户偏好.通过Spark、Flink、Redis等技术实现推荐系统框架的构建,并对深度学习推荐模型进行部署,实现了个性化电影推荐系统.In the era of big data,the traditional collaborative filtering recommendation algorithm has been difficult to meet people's needs for personalized recommendation.To solve this problem,this paper proposes a deep learning recommendation model(GCN-MLP)fused with graph neural network.The model is based on the graph neural network to infer the interactive relationship between the user and the movie,and based on the multi-layer perceptron to mine the complex combination of multiple feature domains,and the two are combined to obtain the corresponding user preferences.The recommendation system framework is constructed through Spark,Flink,Redis and other technologies,and the above-mentioned deep learning recommendation model is deployed,and finally a personalized movie recommendation system is designed and implemented.

关 键 词：图神经网络 深度学习 推荐系统 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]