Name: MuSe Music Sentiment Analysis
Creator: Kaggle
License: https://creativecommons.org/publicdomain/zero/1.0/

About this Dataset

MuSe Music Sentiment Analysis

Music Tags, Metadata, & Audio Features

By [source]

About this dataset

MuSe is proud to present a unique dataset of sentiment information for 90,408 songs. Our data includes user-generated tags from Last.fm, artist and title metadata as well as Spotify and MusicBrainz IDs. This combination of data points allows researchers to explore music across various dimensions such as genre, year, or emotion! With our accompanying Jupyter notebook demonstrating how one can easily generate emotion tags for any song via the Last.fm API, you can use MuSe's dataset to create your own entertaining experiments and explorations! MuSe offers an unparalleled opportunity to uncover the intricate connections between music and emotions in an insightful new way!

More Datasets

For more datasets, click here.

Featured Notebooks

🚨 Your notebook can be here! 🚨!

How to use the dataset

In this guide we'll discuss how you can use this dataset to help power your own analysis and discoveries.

First, let's discuss the columns in the dataset. The columns are track (the name of the song), artist (the name of the artist), valence_tags (tags associated with the valence of the song), arousal_tags (tags associated with the arousal of the song), dominance_tags (tags associated with the dominance of then song). Looking into these tags can give insight into emotion that different people associate with certain songs or artists which is a great place to start for sentiment analysis.

Once you deconstruct who associates what emotions with what songs/artists you can use Spotfy ID and MusicBrainz ID from our dataset to get further information about each song such as its genre or release year on other databases. This can also beA great way to compare different genres as well as types of media to observe any differences or similarities between them when it comes sentiment or association oof emotions towards certain topics in today's culture!

Lastly, if you're looking for some inspiration on how this data could be used we have provided a Jupyter notebook that demonstrates how one could fetch those user-generated lastFM tags from within your own experiments! Have fun exploring our MuSe Dataset!

Research Ideas

Creating personalized music playlists based on user sentiment preferences for arousal, dominance and valence.

Analyzing musical trends over time by looking at the affective tags associated with songs released in particular years or decades.

Developing tools that can generate an affective ‘profile’ of a song by combining the valence, dominance and arousal scores together with additional audio features such as tempo or rhythm

Acknowledgements

If you use this dataset in your research, please credit the original authors.
Data Source

License

License: CC0 1.0 Universal (CC0 1.0) - Public Domain Dedication
No Copyright - You can copy, modify, distribute and perform the work, even for commercial purposes, all without asking permission. See Other Information.

Columns

File: muse_dataset.csv

Column name	Description
track	The title of each song. (String)
artist	The artist of each song. (String)
valence_tags	User-generated mood labels. (String)
arousal_tags	User-generated energy labels. (String)
dominance_tags	User-generated power labels. (String)

Acknowledgements

If you use this dataset in your research, please credit the original authors.
If you use this dataset in your research, please credit .

Tables

Muse Dataset

@kaggle.thedevastator_muse_music_sentiment_analysis.muse_dataset

7.67 MB
90,408 rows
8 columns

CREATE TABLE muse_dataset (
  "id" BIGINT,
  "track" VARCHAR,
  "artist" VARCHAR,
  "valence_tags" DOUBLE,
  "arousal_tags" DOUBLE,
  "dominance_tags" DOUBLE,
  "mbid" VARCHAR,
  "spotify_id" VARCHAR
);