S16K: Multi-Label Music Emotion Recognition Dataset with 16k+ Songs

More than 16,000 songs from NetEase Cloud Music are included in this multi-label music emotion recognition dataset. MFCC features were extracted from the middle 30 seconds of each song using librosa. The dataset was created by joyfuljune and last updated on 2026-01-29.

Use Cases

• Train multi-label emotion classifiers based on MFCC audio features.
• Benchmark music emotion recognition algorithms using a large-scale collection of songs.
• Analyze the relationship between audio features and emotional tags in music.
• Develop music recommendation systems based on predicted emotional content.

Strengths

• Contains more than 16,000 songs, providing a substantial scale for training.
• MFCC features are extracted from a standardized 30-second segment of each song.
• Includes a balanced multi-label subset of 9,822 songs for specific use cases.

Limitations

• Row count for the full dataset is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data may reflect the geographic or cultural bias inherent to the NetEase Cloud Music source.

Provenance

Source

NetEase Cloud Music

Collection Method

MFCC features extracted by librosa from the middle 30 seconds of songs.

Freshness

Last updated 2026-01-29 12:57:32; freshness should be verified.