Learning Network simulation with a contextualized recommender system

Abstract

The main goal of our project is to research how recommender system technologies behave in Learning Networks of different sizes. One of the questions in this research is how to tackle the problem of the difficulty for learners to navigate through the Learning Network: what is the most effective way to reach the learning goal or which learning activity can be studied best after finishing a particular one? The aim of this program is use two different recommendation technologies to recommend learning activities to learners in Learning Networks of different sizes. The performance of the recommendation technologies is explore based on ROC curve analysis (accuracy, precision, recall and F1) and educational measures like effectiveness, efficiency and satisfaction.