摘要
在信息过载时代,从海量信息中寻找感兴趣的信息是一件非常困难的事.推荐系统可以从大数据中挖掘用户的偏好信息,从而向用户提供精确的个性化推荐服务.近年来,深度神经网络在推荐系统中得到了广泛的应用,具有独特的特征提取能力.本文对推荐系统进行梳理,在讨论传统推荐算法的基础上,综述了基于深度神经网络的推荐系统的研究进展,分析了与传统推荐方法的区别与优势,归纳了推荐系统的性能评价指标.介绍了所提出的三个基于深度学习的推荐模型.并对推荐系统的未来发展趋势进行展望.
In the age of information overload,it is very difficult to find information that interests oneself from a large amount of information.To provide users with accurate personalized recommendation services,the recommendation system is used to mine user preference information from big data.In recent years,deep neural networks have been widely used in recommendation system and have excellent feature extraction capabilities.Therefore,this paper first sorts out the concept of recommendation system,and reviews the research progress of recommendation systems based on neural networks,and analyzes their differences and advantages against traditional recommendation methods.Then,the performance evaluation index of the recommendation system is summarized.Finally,we introduce three recommendation models proposed by us and forecast the future development trend of recommendation system.
作者
刘方爱
王倩倩
郝建华
Liu Fangai;Wang Qianqian;Hao Jianhua(School of Information Science and Engineering,Shandong Normal University,250358,Jinan,China;School of Data and Computer Science,Shandong Women′s University,250002,Jinan,China)
出处
《山东师范大学学报（自然科学版）》
2021年第4期325-336,共12页
Journal of Shandong Normal University(Natural Science)
基金
国家自然科学基金资助项目(61772321)
山东省自然科学基金重点项目(ZR202011020044).
关键词
推荐系统
深度神经网络
深度学习
评价指标
注意力机制
recommendation system
deep neural network
deep learning
evaluation metrics
attention mechanism
