EmotionInsight and BeatSync: a music recommendation system based on sentiment analysis

Although tailored music recommendation systems have seen major developments in recent years, their integration with user emotions derived from several input modalities is still rather understudied.EmotionInsight and BeatSync is a new music recommendation system presented in this work that dynamically matches musical suggestions with users inferred emotional states derived from both textual and video-based inputs.Using natural language processing (NLP) methods, the system extracts sentiment and emotional context.Face expressions and emotional cues from video frames are analyzed concurrently using deep learning models.Using a multi-modal sentiment fusion technique, the extracted emotional elements are then mapped to a well-chosen music database, so providing context-aware and emotionally resonant song recommendations.