Generic autonomic model for middlewares of management of eco-connectivist learning environments

This article presents a Generic Autonomic Model based on Multi-Agent Systems, to characterize the reflection capabilities of ARMAGAeco-c, which is an autonomic reflective architecture for the management of eco-connectivist learning environments. In the eco-Connectivism is planed the configuration, the stabilization and the unification of an ecology of knowledge, composed of emerging clusters of personal learning environments. For this, it uses association rules, which are dynamically adapted for the analysis of the connections that occur between apprentices, in a socialized context of learning. The multi-agent system provides an intelligent model of dynamic configuration of roles, with the objective of providing configuration, stabilization and unification of the ecology of eco-connectivist knowledge, using concepts inspired from the ecology, the data mining and the collaborative filtering. Adaptive capacities of personal learning environments are provided by recommendation agents, which reason about emerging clusters based on a model of ecological survival. This enables to characterize the ecological unification of knowledge between individuals.