Characterization and analysis with xAPI based graphs for adaptive interactive learning environments

In e-learning, insights from the analysis of usage tracking data can help improve teaching and learning, e.g., with learning analytics to identify strengths and weaknesses of learners or course material, or for targeted help for individual students. One analysis approach is to examine the graph networks of interaction usages. Adaptive e-learning systems (ALS), which personalize the learning experience to the learners' needs, can make use of relationship information in graph networks to determine the best adaptation strategy. For example, ALS can use graph algorithms to detect central activities that have high influence to the users or to learning objects. This paper shows how to make use of the Experience API (xAPI) protocol and graph networks for its application in adaptive interactive learning environments such as computer simulations and serious games. A prototype implementation hints at the feasibility of the concept and its practical implications.