基于Spark平台的电子商务个性化信息推荐方法  

Personalized information recommendation method of E-commerce based on spark platform

作　　者：李加军 LI Jia-jun(School of Data Science,Guangzhou Huashang College,Guangzhou 511300,China)

机构地区：[1]广州华商学院数据科学学院,广州511300

出　　处：《信息技术》2023年第10期66-71,共6页Information Technology

基　　金：广东财经大学华商学院校内导师制科研项目(2020-HSDS28);广东财经大学华商学院校级质量工程电子商务特色专业(HS2019ZLGC05)。

摘　　要：互联网上数据传播量日益增加,但信息使用率却很低,消耗用户大量精力,针对这个问题,提出一种基于Spark平台的电子商务个性化信息推荐方法。Spark平台通过弹性分布式内存数据集,可将中间计算结果直接保存至内存中,建立用户喜好模型;使用评分机制计算不同个体偏好商品,形成推荐列表;引入挖掘隐含信息的矩阵分解算法,将未知参数转化为已知量,提高个性化信息推荐精准度。仿真对比实验,从用户满意度、信息熵值和运行速度三个角度,验证了所提方法可以实现优质且高效的电子商务个性化信息推荐工作。The amount of data dissemination on the Internet is increasing,but the information utilization rate is very low,which consumes a lot of consumer energy.To solve this problem,an E-commerce personalized information recommendation method based on spark platform is proposed.Spark platform can directly save the intermediate calculation results to memory and establish user preference model through elastic distributed memory data set.Then,the scoring mechanism is adopted to calculate different individual preference products and form a recommendation list.The matrix decomposition algorithm for mining hidden information is introduced to transform unknown parameters into known quantities to improve the accuracy of personalized information recommendation.Simulation and comparison experiments verify that the proposed method can achieve high-quality and efficient E-commerce personalized information recommendation from the perspectives of user satisfaction,information entropy and running speed.

关 键 词：个性化信息推荐 Spark平台 用户喜好模型 评分机制 并行化协同过滤 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]