基于边缘计算的融合多因素的个性化推荐算法  

Personalized recommendation algorithm fusing multiple factors in edge computing

作　　者：金焕章 朱容波[1] 刘浩 陈慧敏 JIN Huanzhang;ZHU Rongbo;LIU Hao;CHEN Huimin(College of Computer Science,South-Central Minzu University,Wuhan 430074,China)

机构地区：[1]中南民族大学计算机科学学院,武汉430074

出　　处：《中南民族大学学报（自然科学版）》2024年第2期217-225,共9页Journal of South-Central University for Nationalities：Natural Science Edition

基　　金：国家重点研发计划资助项目(2020YFC1522600);国家自然科学基金资助项目(62062019);中央高校基本科研业务费专项资金资助项目(CZZ21003)。

摘　　要：针对传统推荐算法使用单一上下文信息不能有效地解决信息过载、数据稀疏、冷启动的问题,提出了一种基于边缘计算的融合多因素的个性化推荐算法F-SVD和新的用户相似度计算方法F-PEARSON(改进后的PEARSON相关系数).在边缘服务器处理个性化用户数据以分散云服务器的压力,目前大多采用的集中式处理方式无法在爆炸性式增长的数据下提供准确的推荐,在云服务器融合多因素挖掘用户之间的潜在关系,从而构建预测F-SVD算法.实验结果表明:在公开数据集MovieLens上,与传统算法相比,所提出的算法在RMSE和MAE上的误差更小,精确度提升了2.2%.As traditional recommendation algorithms using single context information cannot effectively solve information overload,data sparsity,and cold start problems,a personalized recommendation algorithm F-SVD based on edge computing fusing multiple factors and a new user similarity calculation method F-PEARSON(improved PEARSON correlation coefficient)are proposed.Processing personalized user data at the edge server to spread the pressure of the cloud server,the current mostly centralized processing cannot provide accurate recommendations under the explosive growth of data,and fusing multiple factors at the cloud server to mine the potential relationships between users to build the predictive FSVD algorithm.The experimental results show that on the public dataset MovieLens,compared with the traditional algorithm,the proposed algorithm has smaller errors on RMSE and MAE,and the accuracy is improved by 2.2%.

关 键 词：边缘计算 个性化推荐 多因素 数据稀疏 冷启动 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]