AGENP: An ASGrammar-based GENerative Policy Framework

Abstract Generative policies have been proposed as a mechanism to autonomously adapt to changes in a system to achieve the systems's goals with minimum to none human intervention. The combination of generative policies and learning mechanisms can help a coalition system to be more effective when working in a distributed, continuously transforming environment with a diverse set of members, resources, and tasks. Learning mechanisms based on logic e.g., Inductive Logic Programming (ILP), have several properties that make them suitable and attractive for the creation and adaptation of generative policies. Some of the ILP features include representation of the data and problems in a declarative way, the ability to learn a general model with a small number of examples, and with an existing background knowledge. Recently, Answer Set Programming (ASP) have been recognized as a powerful language for knowledge representation and reasoning for which ILP has been successfully applied to. This paper proposes AGENP, a ASGrammar-based GENerative Policy Framework for Autonomous Managed Systems (AMS) that aims to support the creation and evolution of generative policies by leveraging ILP. We describe the framework components, the required inputs, data structures, and mechanisms to support the refinement and instantiation of policies, identification of policy violations, and monitoring of policies and policy adaptation according to changes in the AMS and its context. We also present the main workflow for the global and local refinement of policies and their adaptation based on ASP and ILP concepts and tools, and present example scenarios where the AGENP framework can be useful for the self-generation of policies.
  • Seraphin Calo (IBM US)
  • Irene Manatos (IBM US)
  • Dinesh Verma (IBM US)
  • Mark Law (Imperial)
  • Elisa Bertino (Purdue)
  • Alessandra Russo (Imperial)
Date Sep-2018
Venue 2nd International Workshop on Policy-based Autonomic Data Governance (PADG 2018)