A novel self-organizing multi agent-based approach for multimedia documents adaptation

The development of smart systems using pervasive computing is evolving rapidly due to the increased use of mobile and sensor technologies. These systems usually involve contextual information related to users and devices, which may put constraints on the presentation of multimedia documents. An effective solution consists in adapting their contents according to the current context. Most of existing adaptation approaches try to select, compose and execute adaptation services in order to provide adapted documents. Even so, they, however, do not address the need to deal with dynamic environments for which there does not exist a global view about the available services. Also, they do not care about the increasing number of users' requests. These aspects have a direct influence on the quality of service. To overcome these limitations, we propose a novel self-organizing multi agent based architecture for adapting multimedia documents. The agents cooperate in a decentralized structure to optimize the selection, composition and execution of service plans. The self-organization of the system is ensured by updating the structural relations between the agents according to trust values. To show the feasibility of our proposal, an experimental study is conducted. Overall, the results obtained are encouraging.