基于MyMediaLite平台的推荐方法探究  

Research on the Recommendation Method Based on the MyMediaLite Platform

作　　者：林楠 杨文渊 马伊莉 朱婷婷 陈圣磊[1] 

机构地区：[1]南京审计大学经济与贸易学院,江苏南京211815

出　　处：《软件工程》2018年第1期18-20,11,共4页Software Engineering

基　　金：南京审计大学大学生创新创业训练计划资助(编号:2016AX04001Z)

摘　　要：推荐系统是互联网和电子商务的产物。它是建立在对海量数据训练的基础上的一种智能平台,能够向顾客提供个性化的信息服务和决策。随着电子商务大数据的高速发展,推荐系统正逐渐成为学术界的研究热点之一。针对推荐系统理论性强、内容抽象的特点,本文介绍了以MyMediaLite为平台的个性化推荐实践方案,并详细阐述了其具体的实施过程。通过介绍MyMediaLite的系统结构框架,以及分析基于MyMedia Lite的实验过程,为研究者使用MyMediaLite推荐系统库进行算法研究奠定了基础。The recommendation system is the product of Internet and e-commerce.It is an intelligent platform built on the basis of massive data training.It provides personalized information service and decision-making to customers.With the rapid development of big data in electronic commerce,the recommendation system is becoming one of the hot topics in the academic field.In view of the highly theoretical and abstract nature of the recommendation system,this paper introduces the personalized recommending practice scheme based on My Media Lite,and expounds its specific implementation process in detail.By introducing the system framework of My Media Lite and analyzing the experimental process based on My Media Lite,it establishes a foundation for researchers to conduct studies on the algorithms with My Media Lite recommendation system library.

关 键 词：个性化推荐 机器学习 MyMediaLite 推荐系统 协同过滤 

分 类 号：TP317[自动化与计算机技术—计算机软件与理论]