Presenting information to an e-learning environment is a challenge, mostly, because ofthe hypertextlhypermedia nature and the richness ofthe context and information provides. This paper proposes a dynamic semantic model for e-learning system based on the psycholinguistic theories of human memory, Spreading Activation Network (SAN). This work employs a SAN as a technique to provide the interface's action selection mechanism in an uncertain environment. The paper combines the SAN with the temporal logic to provide an elearning system that a learning activity level evolves according to their expected contextual relevance. The system differsfrom the other e-learning by representing dynamic associations between learning activities and the relevance subjects. This system equipped with an Event-Triggered learning interface (context) adaptation component. This component provides multiple parallel processes for perception. These processes provide context screen selection and learning task...