Combining RPA and AI to recognize learners’ emotions in Cyberspace: Experiment at some Vietnamese universities

This study presents a hybrid system that integrates Robotic Process Automation (RPA) and Artificial Intelligence (AI) to recognize learners’ emotions in Cyberspace. The system automatically collects learner-generated content from educational social media platforms based on RPA technology and applies a deep learning model combining Long Short-Term Memory (LSTM) and Convolutional Neural Networks (CNN) for sentiment classification. A Vietnamese-language dataset, the NEU dataset, consisting of 29,661 manually annotated comments, was curated, reflecting the informal, slang-rich discourse of Gen Z learners. The model categorizes emotions into four classes: Neutral, Positive, Negative, and Strong Negative. Experiments show the model achieves high classification accuracy, particularly in distinguishing strongly negative sentiment. The proposed system supports educational institutions in monitoring and detecting learners’ emotional trends, allowing for timely psychological interventions. This contributes to building increased engagement between learners and schools, creating an enjoyable learning journey based on a learner-centered approach.