Survey on Research Progress of Recommendation Algorithms
LI Meng-hao1,ZHAO Xue-jian1,YU Yun-feng2,SONG Xue-yong2,SUN Zhi-xin1
1(School of Modern Posts,Nanjing University of Posts and Telecommunications,Nanjing 210003,China)2(Guoji Beisheng (Nanjing) Technology Development Co.,Ltd,Nanjing 210001,China)
Abstract:With the development of the Internet,the amount of global data has exploded,and information overload is serious.How to obtain the information you really want has become a long-term problem that plagues people.In this background,recommendation algorithms have been widely used in various fields.This article first introduces the classification and important evaluation indexes of related recommendation algorithms.This article introduces the current research progress of recommendation algorithms in each category,including the basic recommendation principles and research progress of traditional recommendation algorithms and the application of neural networks in recommendation algorithms,and summarizes them in the end.At the same time,common problems of recommendation algorithms such as data sparsity,cold start and scalability are analyzed.Finally,the shortcomings of existing recommendation algorithms and some problems encountered in the application are proposed,and the research hotspots of future recommendation algorithms are introduced.
李孟浩,赵学健,余云峰,宋学永,孙知信. 推荐算法研究进展[J]. 小型微型计算机系统, 2022, 43(3): 544-554.
LI Meng-hao,ZHAO Xue-jian,YU Yun-feng,SONG Xue-yong,SUN Zhi-xin. Survey on Research Progress of Recommendation Algorithms. Journal of Chinese Computer Systems, 2022, 43(3): 544-554.