Spanish Reddit Comments for Suicide and Depression Detection

192,347 rows for training, 33,944 for testing, and 22,630 for validation. It consists of Spanish-translated Reddit comments labeled as suicidal or normal, derived from the 'Suicide and Depression Detection' dataset by Nikhileswar Komati.

Use Cases

• Train a binary text classifier to distinguish between 'suicidal' and 'normal' comment labels.
• Analyze linguistic patterns in Spanish-translated social media text for mental health research.
• Fine-tune a language model on a large corpus of labeled Spanish Reddit comments for sentiment or risk detection.

Strengths

• Large scale with 248,921 total labeled text samples (192,347 training rows).
• Explicitly structured for machine learning with predefined train, test, and validation splits.
• Derived from a known source dataset ('Suicide and Depression Detection' on Kaggle), providing traceable provenance.

Limitations

• Data consists of translations to Spanish, which may introduce linguistic artifacts and not reflect native Spanish online discourse.
• No column details or sample data are provided, limiting understanding of the specific features and label distribution.
• The original data source from Reddit may contain inherent biases in user demographics and self-reporting.

Provenance

Source

Derived from the 'Suicide and Depression Detection' dataset by Nikhileswar Komati on Kaggle.

Collection Method

Reddit comments were translated into Spanish. The original dataset was gathered from social media.

Time Range

null

Freshness

null

Geography

null