Supporting Confidentiality in UML: A Profile for the Decentralized Label Model
Paper i proceeding, 2004

We present a way of incorporating a decentralized label model into the UML by defining a profile which is the built-in extension mechanism of the UML. Our profile permits specifying the confidentiality of data in UML by offering annotations for classes, attributes, operations, values of objects, and parameters of operations. Furthermore, the profile supports the creation of Jif (Java information flow) code. The Jif compiler guarantees that the specified confidentiality constraints are not violated. Our approach is appealing in the sense that it offers the possibility to consider confidentiality in UML and that the obtained code is guaranteed to behave correctly.


Rogardt Heldal

Chalmers, Institutionen för datavetenskap, ProSec

Chalmers, Institutionen för datavetenskap

Steffen Schlager

Jakob Bende

Chalmers, Institutionen för datavetenskap

3rd International Workshop on Critical Systems Development with UML


Data- och informationsvetenskap

Mer information