ABSTRACT
The COVID-19 pandemic has changed the education landscape considerably. In many schools and universities, the traditional in-person teaching and learning was changed into remote online teaching and learning. Some universities tried to utilize various types of online learning platforms, such as Moodle Learning Management System (LMS), to facilitate student learning interaction both synchronously and asynchronously. In the implementation of Moodle, the function of faculty leadership in evaluating and providing feedback on every lecturer's teaching and learning process based on valid evidence to ensure the quality of education is required. This research aims to provide a model for evaluating lecturer performance and clustering student learning activities in online learning environments. The data sample was directly extracted from the Moodle database, which consists of activities from 16,523 courses. The data were analyzed and evaluated by taking a normative criterion, which clusters the course into three categories. The analysis stated that there were 4,814 courses (29.1%) categorized as active, 6,248 courses (37.8%) categorized as less active, and 5,461 courses (33,0%) categorized as inactive. It was also concluded that 11,062 courses (70%) have online learning activities. Furthermore, the course data categorization results were used as the prediction target class. Then, 13.218 course data samples were utilized as training data, and 3.305 were used as testing data. Finally, the Decision Tree Model is applied to create a predictive model. This predictive model has excellent accuracy, sensitivity, and specificity; each scored 100%.
