Policies REfined DYnamically and Kept On Track
PREDYKOT is an European project funded by France in the context of the call 5 of the ITEA 2 Eureka program for the period: 2011-2014. The consortium is composed of 16 partners from 4 countries (France, Finland, Spain, Turkey). The PREDYKOT aims at developing a new approach for managing and adapting their policies on the fly. This approach based on logics and complex events reasoning and complex events processing will enable organizations to respond dynamically to complex events and reconfigure their policies to adapt to the new situations and guarantee the stability of their environments.
The PREDYKOT project shifted the focus of the management of policies from automation of reactive actions and basic controls to a more intelligence-based approach, using critical intelligence to constantly update these policies and consequently further improve business processes. The intelligent mechanisms developed in the project ensure that security policy not only becomes and stays efficient but also incorporates contextual information to enable the policy to be dynamically refined on a continuous basis. The new policy and reasoning languages developed by PREDYKOT combined the best of different policy-specification languages, events from management policies and authority from security policies on the basis of existing standards such as the extensible access control mark-up language, XACML, thereby enabling the specification and interpretation of security policy to be simplified. Reasoning tools were used to analyze policies specified by this language and to detect, for example, possible conflicts or inconsistencies in the policy specification.
A complete suite of software modules have been developed to work together as an ecosystem for monitoring and improving policies and keeping them on track:
Reasoning engines on user activities, policy changes and contextual information;
Smart nodes as actuators or sensors for capturing the context and triggering reactive actions to update the policies;
interfaces with security-information and event-management systems;
Complex events processing and distributed data fusion;
Workflow and security portal to apply feedback on the policy; and
Steering dashboard.
With respect to smart nodes, for example, an ontology with reasoning rules enables the nodes to sense their environment and share knowledge, providing adaptive mechanisms to adjust the policy as the world changes and as new information becomes available. Anticipating the massive deployment of smart nodes in environments like personal mobile devices, PREDYKOT developed models of smart nodes that are able to calculate metrics, adapt policies and provide in-depth accurate information to specialised reasoning engines.
In this project, the contributions of LISSI are centered around the development of ontology based reasoning capabilities of smart noes as well as temporal and spatiotemporal reasoning models for the recognition of situations and analysis policies. The LISSI have expanded the scope of Predykot in order to develop smart nodes that are able to recognize the activities of elderly people at home in the context of cognitive assistance services and adapt the policies controlling the living environment.
