Machine Learning Enhanced Access Control For Big Data
INTERNATIONAL JOURNAL OF COMPUTER SCIENCE AND NETWORK SECURITY(2020)
摘要
Access Controls (AC) are one of the main means of defense in IT systems, unfortunately, Big Data Systems are still lacking in this field, the current well-known ACs are vulnerable and can be compromised because of policy misconfiguration and lack of contextuality. In this article we propose a Machine Learning approach to optimize ABAC (Attribute Based Access Control) with the aim to reduce the attacks that are overlooked by the hardcoded policies (i.e: users abusing their privileges). We use unsupervised learning outlier detection algorithms to detect anomalous user behaviors. The Framework was implemented in Python and its performance tested using the UNSW-NB15 Data Set.
更多查看译文
关键词
Access Control, Big Data, Machine Learning, Outlier Detection, ABAC, Security
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要