A Machine Learning Approach on Multimedia Data for Predicting Student Anxiety

Student anxiety is a critical issue that significantly affects academic performance and overall wellbeing, making early detection essential for timely interventions and appropriate support. This study proposes a machine learning-based system to predict student anxiety by analysing multimedia data (images, videos) and textual data. The system has three input options: users can upload images or videos for analysis or input text, such as social media content, responses to questionnaires. Deep learning models such as Convolutional Neural Networks (CNN), VGG16, AlexNet, and GoogLeNet, analyze emotional cues in images and videos. For textual data, Natural Language Processing (NLP) techniques like XGBoost and TF-IDF, will be used to assess psychological tone and sentiment in text. By integrating insights from both multimedia and text data, the system aims to provide accurate predictions of anxiety levels. This tool is designed for educators and mental health professionals to better assess and predict anxious states in students, enabling early intervention. Ultimately, the project seeks to enhance mental health support within educational environments, fostering a healthier, more supportive learning atmosphere for students.