Using Massive Processing and Mining for Modelling and Decision Making in Online Learning Systems

Abstract

Online Learning and Virtual Campuses have become commonplace paradigms for distance teaching and learning. Unlike face to face teaching and learning methods in which teachers and managers can take decisions based on information from everyday classroom activities, decision making in online learning becomes more complex due to the online setting. Teachers need to get information from the online learning system on the learning processes and learners' activities in order to better support them during the learning process. On the other hand, managers need information on the usage of computational resources of the Virtual Campus to make the computational infrastructure as much efficient as possible. In this work we will address the use of massive processing and data mining techniques to assist teachers, managers and developers of a Virtual Campus in their decision making, aiming to better support teaching and learning processes. Our approach is based on processing log files of the online learning system (Virtual Campus, specific learning platform, document repositories) which keep information on online users during their interaction with and within the system. Log files, which are nowadays commonplace in all learning management systems, tend to be large to very large in size, and thus require a massive processing and then statistical analysis and data mining techniques to extract useful information on user activities, resource usage in the Virtual Campus and web content access, among others.