Cross‑Cultural Corpus of Twelve Non‑Western Musical Traditions: Curated Dataset

This dataset provides a curated, community‑informed track list of 12 non‑Western musical traditions, totalling 363 unique tracks. It serves as the foundational corpus for the research paper "Signal-Based Acoustic Taxonomy of Non-Western Musical Traditions: Cross-Cultural Structure in Pre-Processing Audio Descriptors". Contents: Cross-Cultural-Audio-Corpus-Dataset.csv - Columns: tradition, track_id, track, artist, style, era. Each row represents one track, identified by tradition, artist, title, a descriptive style tag, and an approximate recording era. Traditions represented: Arabic/Maqam, Central Asian, Chinese, Hindi/Hindustani, Indonesian, Japanese, Malayalam, Nepali, Sinhala, Tamil, Tibetan, West African. Tracks span traditional, classical, popular, and contemporary styles, ranging from the early 20th century to the 2020s. Corpus construction: The track list was seeded from musicological literature and refined through community feedback on public forums. Each suggestion was verified and annotated with style and era metadata. Intended use: Reference track list for cross‑cultural MIR research; seed corpus for music classification, recommendation, or similarity models; transparent baseline for acoustic comparisons across world music traditions; resource for within‑tradition cross‑era analysis. Limitations: This dataset contains metadata only; it does not include audio files. Era assignments are approximate, and recording quality varies across the corpus.