【摘要】 随着大数据时代的到来,大数据机器学习成为了重要的研究热点。电商网站越来越重视对于用户推荐系统的搭建,然而推荐系统对于要过滤的信息的规模往往是巨大的,Spark平台作为新一代计算框架,具有对大数据处理能力强等特点,大大提高了推荐系统的运行效率。本文基于Spark平台的以上特点,提出并设计了基于Spark平台的电商推荐系统,设计和实现了协同过滤算法,提高了系统的运行效率。实验结果表明基于Spark平台的推荐系统运行效率是Hadoop平台的8倍以上,增强了电商网站的用户体验。更多还原

【Abstract】 With the advent of large data age,large data machine learning has become an important research hotspot.Electric business website are more and more attention to build the user recommended system.Spark platform,as a new generation of computing framework,with a large data processing capabilities and other characteristics,greatly improving the recommended system operating efficiency.This paper is based on the domestic and international research on the Spark platform,proposed and designed based on the Spark platform for electricity providers recommended system,and achieve the recommended algorithm of parallel computing,enhance the user experience of the e-commerce site.更多还原