一种面向Hadoop平台的自适应授权机制  

An adaptive authorization scheme of Hadoop platform

作　　者：左冉 赵刚[1] ZUO Ran;ZHAO Gang(School of Information Management,Beijing Information Science&Technology University,Beijing 100192,China)

机构地区：[1]北京信息科技大学信息管理学院,北京100192

出　　处：《北京信息科技大学学报（自然科学版）》2021年第1期6-10,共5页Journal of Beijing Information Science and Technology University

基　　金：国家自然科学基金资助项目(71701020)。

摘　　要：针对Hadoop授权系统对系统管理员强依赖且不够灵活的问题,以增强授权工作的时效性与全面性为目的建立了一种自适应授权机制。考虑到用户访问行为随时间动态变化特点,提出通过用户的审计日志数据进行用户行为分析的方法,对用户可疑状态进行评估,为Hadoop平台提出较为准确的权限调整建议。建立的面向Hadoop平台的自适应授权机制,能够根据用户实时行为数据动态调整用户对Hadoop平台相关资源的访问权限。Hadoop平台的授权工作是保护大数据环境下用户敏感信息和隐私数据的重要工作,通过建立一种自适应授权机制,能够为Hadoop平台的授权工作提供助力。For the strong dependence on system administrators and insufficient flexibility of Hadoop authorization system’s,an adaptive authorization scheme is established for the purpose of enhancing the timeliness and comprehensiveness of authorization work.Considering the dynamic changes of user access behavior over time,a method of user behavior analysis based on user audit log data is proposed,the suspicious state of users is evaluated,and suggestions of permission adjustment are provided for the Hadoop platform.The established adaptive authorization scheme oriented to the Hadoop platform can dynamically adjust the user’s access rights to related resources on the Hadoop platform based on the user’s real-time behavior data.The authorization work of the Hadoop platform is important for the protection of the sensitive information and private data of users in the big data environment.By establishing an adaptive authorization scheme,the authorization of the Hadoop platform can be facilitated.

关 键 词：用户行为分析 Hadoop授权 授权机制 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构]