期刊文献+

基于Spark的Hybrid推荐算法的研究与实现

Research and implementation of Hybrid recommendation algorithm based on Spark
收藏 分享 导出
摘要 协同过滤推荐技术作为推荐系统的一个重要分支,成为目前应用最广泛的一种推荐算法。但协同过滤算法仍然显露出数据稀疏性问题、可扩展性问题等。为解决上述问题,文章提出了基于Spark平台的基于ALS算法和物品相似度相结合的混合协同过滤算法。本文算法在一定程度上解决了因数据量不足带来的数据稀疏性问题,基于Spark分布式并行计算框架技术也解决了可扩展性问题,同时又提高了算法推荐的准确性。基于Movielens数据集的实验表明,文章算法具有可扩展性高、响应时间短以及推荐精度高等特点。 Collaborative filtering recommendation technology, as an important branch of recommendation system, has become the most widely used recommendation algorithm at present. But the collaborative filtering algorithm still shows the problem of data sparsity, scalability and so on. To solve the above problems, based on the Spark platform, a hybrid collaborative filtering algorithm based on ALS algorithm and item similarity is proposed in this paper. This algorithm solves the problem of sparse number caused by insufficient data, and the distributed parallel computing framework based on Spark also solves the scalability problem, and also improves the accuracy of the algorithm recommendation. Experiments based on Movielens dataset show that the algorithm has the characteristics of high scalability, short response time and high recommendation accuracy.
作者 祝永志 Zhu Yong-zhi(School of Information Science and Engineering,Qufu Normal University,Rizbao 276826,China)
出处 《电子技术(上海)》 2018年第12期59-62,共4页
基金 山东省自然科学基金(ZR2013FL015) 山东省研究生教育创新资助计划(SDYY12060).
关键词 Hybrid推荐算法 ALS SPARK 可扩放性 协同过滤. Hybrid recommendation algorithm ALS Spark scalability Collaborative Filtering
作者简介 通讯作者:祝永志,教授,rizhaozyz@126.com。
  • 相关文献

参考文献8

二级参考文献48

  • 1刘鲁,任晓丽.推荐系统研究进展及展望[J].信息系统学报,2008,0(1):82-90. 被引量:32
  • 2周军锋,汤显,郭景峰.一种优化的协同过滤推荐算法[J].计算机研究与发展,2004,41(10):1842-1847. 被引量:87
  • 3BARALIS E, GARZA P. Item selection for associative classification [ J]. International Journal of Intelligent Systems,2012, 27(3 ): 279- 299. 被引量:1
  • 4YE Jun. Cosine similarity measures for intuitionistic fuzzy sets and their applications [ J]. Mathematical and Computer Modelling, 2011,53(1-2) : 91-97. 被引量:1
  • 5GONG Song-jie, YE Hong-wu. Joining user clustering and item based collaborative filtering in personalized recommendation services [ C ]// Proc of International Conference on Industrial and Information Sys- tems. Washington DC : IEEE Computer Society,2009 : 149-151. 被引量:1
  • 6杨世铭,陶文铨.传热学[M].西安:西北工业大学出版社,2012:28-36. 被引量:1
  • 7JIA Rong-fei,JIN Mao-zhong, LIU Chao. A new clustering method for collaborative filtering[ C ]//Proc of International Conference on Net- working and Information Technology. 2010 : 488- 492. 被引量:1
  • 8Zhao ZD,Shang MS.User-based collaborative-filtering recommendation algorithms on Hadoop.3rd International Conference on Knowledge Discovery and Data Mining.IEEE.2010.478-481. 被引量:1
  • 9Jiang J,Lu J,Zhang G,et al.Scaling-up item-based collaborative filtering recommendation algorithm based on Hadoop.Proc.of the 2011 IEEE World Congress on Services (SERVICES '11).IEEE.2011.490-497. 被引量:1
  • 10Schelter S,Boden C,Markl V.Scalable similarity-based neighborhood methods with mapreduce.Proc.of the sixth ACM Conference on Recommender Systems.ACM.2012.163-170. 被引量:1

共引文献21

投稿分析

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部 意见反馈