Enabling Integrated Data Mining Analysis in Moodle With FlexEDM

Data mining and learning analytics promises to provide educators with valuable insights to improve student learning. Tools allow non-expert investigators to harness the potential of data mining within their respective areas of interest. However, the existing data mining tools are often complex, they are not targeted to educators with different competencies and require learning new environments and skills to benefit from their use. In this paper, we utilize the FlexEDM framework to create a data mining tool that integrates and performs analyses within the Moodle learning management system (LMS). This paper builds upon previous work of the authors who proposed the framework, FlexEDM, which allows developers to create a data mining environment that is flexible to the level of knowledge and skill of the educator. The tool, developed as a plugin to Moodle, assesses students' interaction data generated by the LMS to understand underlying trends and forecast performance. We highlight the results produced by utilizing unsupervised learning to discover trends within a programming course offered through the Moodle learning management system. These trends include; the level of engagement and how quizzes and assignments performance can group students. The tool uses supervised learning to forecast students' final examination performance based on previous or similar course offerings within the LMS selected by the instructor. The work presented illustrates the value of the FlexEDM framework for providing an integrated approach to data analytics. This illustration highlights important considerations for researchers and developers to increase the availability of analytical tools within the constraints of the LMS. It highlights how a guided step-by-step process can enable non-technical educators to perform data mining analysis. Furthermore, it proposes a method for enabling supervised learning by allowing the instructor to select and configure cached models from previous courses in the LMS, for use as training data.