Monitoring learning activities in PLE using semantic modelling of learner behaviour

Abstract

We report on the reflection of learning activities and revealing hidden information based on tracked user behaviour in our widget based PLE (Personal Learning Environment) at Graz University of Technology. Our reference data set includes information of more then 4000 active learners for a period of around two years. We have modelled activity and usage traces using domain specific ontologies like Activity Ontology and Learning Context Ontology from the IntelLEO EU project. Generally we distinguish three different metrics: user centric, learning object (widget) centric and activity centric. We have used Semantic Web query languages like SPARQL and representation formats like RDF to implement a human and machine readable web service along with a learning analytics dashboard for metrics visualization. The results offer a quick overview of learning habits, preferred set-ups of learning objects (widgets) and overall reflection of usages and activity dynamics in the PLE platform over time. The architecture delivers insights for intervening and recommending as closure of a learning analytics cycle[1] to optimize confidence in the PLE. © 2013 Springer-Verlag.