Paper Title
LSTM-BASED MODEL FOR PREDICTING ACADEMIC PERFORMANCE THROUGH ANALYSIS OF EMOTIONS AND PARTICIPATION IN ONLINE LEARNING

Abstract
In the realm of online education, understanding the intricate interplay between emotional states, active participation (measured by the number of posted discussions, replies, and connection time), and academic performance is essential for enhancing the learning experience. This study leverages machine learning al-gorithms and LSTM (Long Short-Term Memory) networks to forecast online ac-ademic performance through the analysis of emotions and social engagement. Re-sults obtained from a sample of student-teachers demonstrate that multiple regres-sion is the most effective method for predicting academic success. This research emphasizes the importance of factoring emotional and social aspects in evaluating online performance and opens new avenues for personalizing learning pathways. Keywords - Emotional state analysis, Natural language processing, Learner performance, Multiple regression, Online learning.