satire detection

The SatirA 2025 dataset consists of a collection of audio segments (approximately 25 hours) and their transcriptions, extracted from satirical programs and news broadcasts in Spanish and annotated as satire or non-satire.