基于Mahout的图书推荐引擎的研究与设计

摘要

The recommendation engine make predictions based on the interest features and operational behavior of users, it recommend to user information of interest or commodities. This paper analyzes the existing mainstream technology and Mahout collaborative filtering algorithm, and an improved recommended logic and an optimization ranking were proposed. The paper also used the distributed storage and parallel computing to analyze and design a recommendation engine for the E-commerce web site' massive books, thus providing users with fast and accurate recommendation service of books.%推荐引擎是一种根据用户的兴趣特点和操作行为做出预测，向用户推荐用户感兴趣的信息或商品的应用引擎。本文分析了现有的主流技术和Mahout协同过滤算法，提出推荐逻辑的改进和排名的优化。还采用分布式存储和并行计算等技术，分析并设计了一个针对电子商务网站中海量的图书的推荐引擎，为用户提供针对电子商务网站中图书的快速、准确的推荐服务。