Improving Learning Analytics from Open-Source Software Data Logs Using Machine Learning and Process Mining Techniques
摘要
Learning analytics plays a crucial role in understanding how learning occurs and how skills are transmitted in contextual settings. Research indicates that participants in open-source environments benefit from and contribute to knowledge generation and skill transfer. In a prior study, we examined these environments, focussing on mailing archives to elucidate the initial stages of learning processes. This current study addresses the subsequent Progression Phase, applying Natural Language Processing (NLP) techniques to analyse the evolution of skill generation and transmission through messages in the email archives of the OpenStack data repositories. By utilizing NLP for tagging and quantifying discussions, we mapped the progression of learning through event logs for each conversation. Our findings demonstrate that this method produces richer logs, allowing events to be interpreted as traces of learning activities and visually represented as network workflows. Additionally, machine learning algorithms are vital in identifying relevant conversations that indicate the occurrence of learning activities.