... The classification of music can be a difficult task since the emotional reaction between listeners can be fairly different for a given song. Much of the current organization for songs is based on an artist's overall genre, rather than on the feeling generated by a song. Attempting to categorize music through engineering techniques is challenging, but can potentially help to minimize these discrepancies between listeners in the sorting process. Identifying the mood of a piece automatically would be extremely useful for sorting large collections of digital music such as those of iTunes or Spotify. The mood of a piece could also improve upon algorithms for identifying similar songs for online radio services like Pandora, basing the similarities on the song's mood rather than on similar artists. Breaking a song down into quantifiable musical components such as rhythm, harmony, and timbre can allow for the matching of songs to specific categories based upon expected data for each type of mood.
See post on Music Stack Exchange:
Some very famous pieces of music have started with a mood description written on the sheet music by the composer.
Some other have become the perfect example of a given mood (Appassionata, Pathetic for Beethoven)
... Classifying music is a very active research subject, especially these days for commercial reasons. Every internet music platform wants to propose and sell tracks to their users by different recommendation systems. Classifying (if possible automatically) music tracks becomes a way to feed such a recommendation system.
If people work with clusters, it is because they encountered difficulties in predicting reliably appartenance of music to finer divisions.
To refine these mood clusters, what you can do is take all the adjectives in each cluster, their synonyms and try to see if you can either group some of them or draw limits between them based on simple concepts.
This is an interesting classification dataset from MIREX (Music Information Retrieval Evaluation eXchange), based on moods. See vocabulary clusters/tags from 2019:
G12 calm, comfort, quiet, serene, mellow, chill out, calm down, calming, chillout, comforting, content, cool down, mellow music, mellow rock, peace of mind, quietness, relaxation, serenity, solace, soothe, soothing, still, tranquil, tranquility, tranquility G15 sad, sadness, unhappy, melancholic, melancholy, feeling sad, mood: sad - slightly, sad song 8 G5 happy, happiness, happy songs, happy music, glad, mood: happy G32 romantic, romantic music G2 upbeat, gleeful, high spirits, zest, enthusiastic, buoyancy, elation, mood: upbeat G16 depressed, blue, dark, depressive, dreary, gloom, darkness, depress, depression, depressing, gloomy G28 anger, angry, choleric, fury, outraged, rage, angry music G17 grief, heartbreak, mournful, sorrow, sorry, doleful, heartache, heartbreaking, heartsick, lachrymose, mourning, plaintive, regret, sorrowful G14 dreamy G6 cheerful, cheer up, festive, jolly, jovial, merry, cheer, cheering, cheery, get happy, rejoice, songs that are cheerful, sunny G8 brooding, contemplative, meditative, reflective, broody, pensive, pondering, wistful G29 aggression, aggressive G25 angst, anxiety, anxious, jumpy, nervous, angsty G9 confident, encouraging, encouragement, optimism, optimistic G7 desire, hope, hopeful, mood: hopeful G11 earnest, heartfelt G31 pessimism, cynical, pessimistic, weltschmerz, cynical/sarcastic G1 excitement, exciting, exhilarating, thrill, ardor, stimulating, thrilling, titillating