基于Storm的实时产品推荐系统研究
首发时间:2016-04-13
摘要:互联网技术的飞速发展推动了电子商务的兴起,越来越多的人们选择通过网络购物,因此向用户推荐适当的产品,能够有效促进用户消费。由于受众极广,购物网站每时每刻都能产生大量的用户行为数据,作为后台处理程序,需要实时高效地处理这些信息,并根据适当的算法,向用户推荐产品。本文针对大数据实时处理的需求,提出了一套基于Kafka消息队列系统、Storm流式计算框架和基于Codis的Redis集群的详细解决方案,并对系统各个组件以及计算方法做了详细说明。最后通过实验验证了系统高效性与有效性,本文对研究实时的大数据处理有着重要的参考意义。
关键词: 实时计算; 产品推荐; Storm; Kafka; Redis; Codis
For information in English, please click here
The Research of Real-time Product Recommendation System Based on Storm
Abstract:With the fast development of internet technology, the E-commerce blooms quickly and an increasing number of people are prone to shop on the Internet. If proper products can be recommened to customers, shopping websites may make more profit. Because of innumerable audiences, huge data of customer activities would be generateed every second. It requires that backend system can handle data of customer actives quickly and efficiently and recommend products to customers accroding to proper algorithms. Inorder to solve this problem, this paper gives a soultion based on Kafka notification queen system, Storm computing framewok and Redis cluster based on Codis. Then every component of the system and algorithm are described in detial. Finally, the efficiency and applicability of the system are verified by experiment. This paper has significant value of resarch in big data processing.
Keywords: Real time computation Recommend proucts Storm Kafka Redis Codis
基金:
论文图表:
引用
No.4680972113695814****
同行评议
共计0人参与
勘误表
基于Storm的实时产品推荐系统研究
评论
全部评论0/1000