Transformer Model in Music Mood Classification
T Thoyyibah. · Zenodo (CERN European Organization for Nuclear Research) · 2022
In recent years, artificial intelligence technology has developed rapidly and has become an inseparable part of life. Artificial intelligence has driven many activities in various sub-fields using machine learning. One of them is in the field of music. Every human being loves music. Music has become a part of human life. Music provides its own atmosphere to the listener which is called mood. There are so many elements of music that can be processed in machine learning, such as lyrics and audio. Lyrics and audio are saved into a music dataset. The author uses 1181 data on Indonesian music in the era of the 70s and 80s. The dataset only takes the refrain. The author uses the Transformer model where the lyrics management uses word embedding and audio processing uses the chromagram feature.