Turkish Music Emotion Dataset from UCI

UCI hosts the Turkish Music Emotion dataset, which contains audio recordings and extracted features for emotion analysis. The dataset's specific size, creator, and creation date are not provided in the available metadata. It is designed for computational analysis of emotional content in music.

Use Cases

• Train a classifier to predict emotion labels from extracted audio features like MFCCs or spectral contrast.
• Analyze the correlation between musical attributes such as tempo, key, and mode and perceived emotional valence.
• Develop models for cross-cultural emotion recognition by comparing features from this Turkish dataset with other music emotion corpora.
• Perform feature engineering and selection to identify which audio descriptors are most predictive for arousal and dominance dimensions in Turkish music.

Strengths

• Dataset is hosted by the reputable UCI Machine Learning Repository, ensuring a degree of curation and stability.
• Focus on Turkish music provides data for a specific cultural and linguistic context often underrepresented in MIR research.

Limitations

• The exact number of audio samples, features, and rows is unknown, making it difficult to assess scale and suitability for specific models.
• Without column details, the specific feature set, emotion label schema, and potential class balance are unclear.

Provenance

Source

UCI Machine Learning Repository

Geography

Presumably Turkey, based on the title.