In the digital era, content consumption has surged with the availability of thousands of movies across various platforms. However, traditional movie recommendation systems largely depend on user ratings, search history, and popularity metrics, often neglecting the most immediate and influential factor — user emotion. This project introduces a novel, real-time Emotion-Based Movie Recommendation System that dynamically understands and responds to a user's current emotional state using facial expression analysis. The proposed system leverages computer vision and deep learning techniques to detect and classify facial emotions using a webcam feed. A pre-trained Convolutional Neural Network (CNN) is used to identify primary emotional categories such as happiness, sadness, anger, fear, disgust, surprise, and neutrality. Once an emotion is identified, the system maps it to a corresponding movie genre—for example, recommending comedies to happy users or inspirational dramas to sad users. To ensure relevance and freshness, the system employs web scraping to fetch live movie data from IMDb, avoiding reliance on static or outdated databases. This integration enables real-time generation of personalized movie suggestions that resonate with the user’s current emotional context. Extensive testing of the system has demonstrated accurate emotion classification and high user satisfaction with the relevance of the recommended content. The implementation showcases the potential of emotionally intelligent systems to transform digital interaction, making user experience more empathetic, engaging, and personalized. This emotion-aware recommender not only improves entertainment discovery but also opens pathways for applications in mental health, education, and human-computer interaction, where understanding emotional context is critical.
Volume 17 | Issue 4
Pages: 26-33