CAJ | 학술논문

互联网技术的发展把电子商务带入了海量数据时代，这种“信息过载”现象不但没有给信息检索带来便利，反而带来了不少困惑。个性化推荐系统正是解决这一问题的有效手段之一。传统的研究主要侧重推荐算法等理论研究，并没有考虑到实际系统的开发难度问题。本文从实际应用出发结合开源框架Mahout来设计与实现推荐系统，首先对Mahout源码相关知识进行分析，研究Mahout推荐引擎的源码实现以及使用方式，然后采用MVC设计模式设计了整个系统架构和平台开发的技术路线，最后结合javaEE技术实现了整个系统功能并进行了实验分析。
호련망기술적발전파전자상무대입료해량수거시대，저충“신식과재”현상불단몰유급신식검색대래편리，반이대래료불소곤혹。개성화추천계통정시해결저일문제적유효수단지일。전통적연구주요측중추천산법등이론연구，병몰유고필도실제계통적개발난도문제。본문종실제응용출발결합개원광가Mahout래설계여실현추천계통，수선대Mahout원마상관지식진행분석，연구Mahout추천인경적원마실현이급사용방식，연후채용MVC설계모식설계료정개계통가구화평태개발적기술로선，최후결합javaEE기술실현료정개계통공능병진행료실험분석。
E-commerce has come into the era of massive data with the development of Internet technology, this"information over?load"phenomenon not only failed to bring convenience to the information of find, but also brings a lot of confusion. Personalized recommendation system is one of effective means to solve this problem, traditional research mainly focuses on the recommendation algorithm research, does not take the difficulty of the problem of the actual system development into account. In this paper, design and implement a recommendation system with Mahout open source framework. Firstly, analysis Mahout with open source, study on the implementation and use of the recommendation engine. Secondly using the MVC design pattern to design the entire system and design the technology road map for development system. Finally, implementation the entire system with javaEE technology, and analysis of experiments based on open data sets .