Abstract: The scheme in this paper is a set of data security monitoring solutions for government data leakage risk based on Internet export network traffic, using big data, machine learning, sensitive data identification, abnormal behavior analysis and other technologies, and combining with government affairs and their characteristics. The scheme realizes government data identification, data flow monitoring, user operation anomaly and interface transmission anomaly monitoring, and realizes active and passive identification, monitoring and early warning of risks in the process of government sensitive data flow.

Key words:

data classification, 
																																							   data security, 
																																							   leakage monitoring, 
																																							   behavioral analysis, 
																																							   retrospective evidence collection, 
																																							   interface safety, 
																																							   safety audit

摘要： 方案是基于互联网出口网络流量，利用大数据、机器学习、敏感数据识别、异常行为分析等技术，结合政务业务及其特征，构建的一套面向政务数据泄露风险的数据安全监测方案.该方案实现政务数据识别、数据流动监测、用户操作异常和接口传输异常监测，实现对政务敏感数据流动过程中所存在风险的主被动识别与监测预警.

关键词: 数据分类分级, 数据安全, 泄露监测, 行为分析, 回溯取证, 接口安全, 安全审计