Journal of Frontiers of Computer Science and Technology ›› 2015, Vol. 9 ›› Issue (5): 513-525.DOI: 10.3778/j.issn.1673-9418.1412023

Previous Articles     Next Articles

Recommender System: Up to Now

ZHU Yangyong1,2, SUN Jing1,2+   

  1. 1. Shanghai Key Laboratory of Data Science, Fudan University, Shanghai 201203, China
    2. School of Computer Science, Fudan University, Shanghai 201203, China
  • Online:2015-05-01 Published:2015-05-06

推荐系统研究进展

朱扬勇1,2,孙  婧1,2+   

  1. 1. 上海市数据科学重点实验室(复旦大学),上海 201203
    2. 复旦大学 计算机科学技术学院,上海 201203

Abstract: Recommender system is the product of cyber age today. There have been many achievements in research and application. This paper makes a comprehensive survey of the recommender system. It proposes three research phases, and points out the milestone events in each stage of recommender system development. In the age of big data, exploiting recommendation in the perspective of data, this paper classifies the recommender system into seven main classes according to the different data used in recommendation, and analyzes and comments the recommended models used in each classification and their advantages and disadvantages. Exploiting big data in the perspective of recommendation, this paper proposes that making recommendation based on big data is one of the promising research directions. Finally, this paper compares the evaluation metrics of recommendation, and gives future research directions.

Key words: recommender system, personalization, collaborative filtering, big data

摘要: 推荐系统(recommender system,RS)是当今网络时代的产物,在技术研究和应用方面取得了很多成果。综述了推荐系统领域的研究状况和进展,提出了3个研究阶段,并指出了每个阶段标志性意义的事件。在当前大数据环境下,从数据的角度看推荐,提出了推荐系统新的分类方法,即根据推荐时所使用的数据不同分为7种类别,同时指出了每个类别使用了哪些推荐模型及其优缺点。提出了在大数据环境下进行推荐是未来推荐系统研究的一个大方向,分析了推荐视角下的大数据机制。最后比较和总结了推荐系统的评价指标,给出了未来的主要研究方向和可能的突破点。

关键词: 推荐系统, 个性化, 协同过滤, 大数据