This paper proposes a methodology for analyzing digital footprints in the Moodle Learning Management System (LMS) to identify procrastination patterns and classify students based on temporal activity dynamics. The study introduces novel metrics, including last_week_intensity (activity concentration in the final week), late_activity_ratio (proportion of activity in the last 25% of the course), and activity_trend (linear regression slope of weekly engagement), to quantify behaviors. Using data from 19 students enrolled in a 17-week Human Resource Management course, the authors apply the K-Means clustering algorithm to categorize learners into three distinct groups: procrastinators, achievers, and diligent underachievers. The results reveal significant correlations between procrastination metrics, engagement duration, with procrastinators demonstrating short-term participation (5 weeks) and high end-of-course activity (88% intensity). The study highlights the potential of LMS data for early identification of at-risk students and designing adaptive pedagogical interventions.