A Multimodal Approach to Emotion-Aware Music Recommendation: Integrating Demographic and Emotional Features with Machine Learning
Main Article Content
Abstract
Music has been considered to be a universal language, which has the ability to affect the feelings and make cultural identities. With the ongoing development of digital technologies, personal music recommendation systems are now fully embedded and become a part of life as the engine provides custom-crafted playlists, which match personal taste. Although conventional music recommendation systems mostly use any previous listening data, the addition of demographic characteristics like age or gender gives a more detailed analysis of the unlike preferences of the user. Also, feelings cues such as facial expressions or body reactions complement the personalization of music recommendations. The paper suggests a multimodal model of recommending music that would be based on both demographic aspects (ages, gender) and emotional indicators (mood, facial expressions) based on the latest machine learning methods. The system is based on the Convolutional Neural Networks (CNNs) and other deep learning models to identify emotions and demographic features to improve the accuracy and relevance of music proposals. The combination of the modalities makes the system provide an opportunity to have a personalized experience in getting exposed to music, which is emotionally stirring as well as contextually suitable. The findings prove that emotion-sensitive music suggestions can greatly enhance user-interaction and satisfaction, resulting to a more personalized and influential music playlists.