| 基于Mahout分布式协同过滤推荐算法分析与实现 |
 点此下载全文 |
| 引用本文：曾志浩，张琼林，姚贝，孙琪.基于Mahout分布式协同过滤推荐算法分析与实现[J].计算技术与自动化,2015,(3):67-72 |
| 摘要点击次数: 1747 |
| 全文下载次数: 75 |
|
|
| 中文摘要:随着信息技术和互联网的发展，在信息过载的时代，用户面对海量的信息，难以正确选择。协同过滤推荐是个性化推荐中比较成熟的算法，但其稀疏性、冷启动、可扩展性问题仍然存在，尤其是不能应用于分布式推荐。在Hadoop平台上，Mahout实现了分布式基于项目的协同过滤推荐算法，该算法能够有效解决传统算法的海量数据处理的效率问题和可扩展性问题。实验结果表明，Mahout上基于项目的协同过滤推荐算法具有较好的计算高效性和可扩展性。 |
| 中文关键词:分布式协同过滤 Mahout 推荐系统 |
| |
| Analysis and Implementation of Distributed Collaborative Filtering Recommendation Algorithm Based on Mahout |
|
|
| Abstract:With the development of information technology and Internet, facing the vast amount of information, it is difficult to correctly choose for users in the era of information overload. Collaborative filtering is a relatively mature algorithm in personalized recommendation, but its sparsity, cold start, scalability problems still exist, especially can not be applied to distributed recommendation. On the platform of Hadoop, Mahout realized the distributed item-based collaborative filtering recommendation algorithm, and the algorithm can effectively solve massive data processing efficiency and scalability problem of the traditional algorithm. The experimental results show that, collaborative filtering algorithm has the high calculation efficiency and good scalability based on Mahout. |
| keywords:distributed collaborative filtering mahout recommendation system |
| 查看全文 查看/发表评论 下载pdf阅读器 |
位访问者,今日访问
