Fayol: um sistema multi-agente de gerência e controle de acesso à plataforma Milos

Abstract

The increase use of learning objects and content platforms that perform the functions of storage, creation, modification and query of these objects on a managed and controlled manner, creates the need for mechanisms to control the access to these platforms. Inserted into a specific content platform that supports the complete life-cycle of learning objects, the present work aims to develop a tool that will provide authentication and authorization (access control) services, integrated and compatible with the technologies already in use on platform. The work makes use of ontologies, semantic web and intelligent software agents technologies that comprise the operating environment of the platform, looking to exploit the potential of these technologies and tools, and its effectiveness in the specification and detailing models and mechanisms for authentication and access control. The analysis of the state of the art shows that the application of multi-agent systems and ontologies technologies on authorization and access control questions is an important, but very recent, research trend. Thus, this work aims to contribute to the advancement of such research. This dissertation proposes an complete ontological model for authentication and access control plus an agent-based, federated and secure communication mechanism that implements this model. As a result, this dissertation explores possibilities not yet considered in this new approach, but present in more traditional mechanismos of authentication and authorization. The evaluation of the proposed model and the implemented mechanism was conducted through functional and performance laboratory experiments, following the practice of evaluation of existing mechanisms for authentication and authorization.