基于属性关联的客体聚合信息级别推演方法  

A Level Inference Method for Aggregated Information of Objects Based on Associated Attributes

作　　者：曹利峰[1] 陈性元[1] 杜学绘[1] 邵婧[1] 

机构地区：[1]解放军信息工程大学,河南郑州450004

出　　处：《电子学报》2013年第7期1442-1447,共6页Acta Electronica Sinica

基　　金：国家973计划前期研究专项(No.2011CB311801);河南省杰出科技创新人才计划(No.114200510001h)

摘　　要：为解决客体关联性引起的泄密问题,本文对客体关联性进行了深入分析,提出了基于属性关联的客体聚合信息级别推演方法.该方法根据客体属性依赖关系,挖掘出高关联度的客体,通过客体关联属性级别模糊集可能性测度,推演出关联客体推导出更高级别信息的可能性,以此指导多级安全网络访问控制策略的制定,控制主体对关联客体的访问,降低系统失泄密的风险.A level inference method for aggregated information of objects based on associated attributes is put forward in order to give a solution to information leakage caused by the association among objects.According to dependency relationships,highly associated objects are found,and then the probability of higher level information inferred by aggregation of associated objects is computed by probability estimates of fuzzy sets on secure level of attribute.The method may contribute to establish access control policy in multi-level secure network,and control restricted access of associated objects in order to reduce the risk on information system.

关 键 词：多级安全 等级保护 客体聚合 推理通道 级别推演 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]