KMi Seminars
Using Formal Methods for the Analysis and Refinement of Policies
This event took place on Monday 24 April 2006 at 10:00

 
Dr. Arosha Bandara Imperial College London

Policy-based approaches to systems management are gaining widespread interest because they allow the separation of the rules that govern the behavioural choices of a system from the functionality provided by that system. In order to deploy policy based management in enterprise systems, it is important to be able to analyse policies to ensure consistency and to ensure that key properties are preserved in the configuration of a managed system. Although many policy languages have been proposed in the literature, the need for policy analysis techniques has not been adequately addressed. Additionally, it is important that system administrators have support for transforming the high-level goals they wish to achieve into concrete policies that can be enforced by the system. This process is called policy refinement and is another area of policy based systems research had been largely ignored in previous work.

In this seminar I will present my research on developing an integrated approach to policy analysis and refinement that is based on the application of formal reasoning techniques. Recognising the importance of being able to analyse policies in the presence of constraints that control their applicability, the method uses a formal representation of policies and managed systems, together with abductive reasoning techniques, to not only detect the presence of conflicts but generate explanations of the circumstances in which a conflict might arise. Additionally, the approach supports policy refinement based on goal elaboration techniques developed for requirements engineering. Since the refinement process requires expertise of the application domain, the technique uses the idea of refinement patterns to allow refinement results to be reused by non-expert users.

I conclude by discussing my ongoing research on the use of other formal reasoning frameworks, such as argumentation and preference logic, to provide enhanced analysis and refinement capabilities.

Download PowerPoint presentation (24.5Mb ZIP file)

 
KMi Seminars Event | SSSW 2013, The 10th Summer School on Ontology Engineering and the Semantic Web Journal | 25 years of knowledge acquisition
 

New Media Systems is...


Our New Media Systems research theme aims to show how new media devices, standards, architectures and concepts can change the nature of learning.

Our work involves the development of short life-cycle working prototypes of innovative technologies or concepts that we believe will influence the future of open learning within a 3-5 year timescale. Each new media concept is built into a working prototype of how the innovation may change a target community. The working prototypes are all available (in some form) from this website.

Our prototypes themselves are not designed solely for traditional Open Learning, but include a remit to show how that innovation can and will change learning at all levels and in all forms; in education, at work and play.