---
license: apache-2.0
task_categories:
  - audio-classification
language:
  - tr
tags:
  - audio
  - turkish
  - türkçe
  - podcast
size_categories:
  - 10K<n<100K
---

# Turkish Podcast RSS Links

This dataset contains **podcast RSS feed metadata** for podcasts classified as Turkish by **both** lingua-py and fast-langdetect. It includes only feed metadata and URLs (e.g., RSS URLs), **not** any podcast audio or episode content.

# Why and how was it built?
I wanted to create this dataset because I could not find an existing one that met my needs. I then downloaded the Podcast Index dataset, which contains 4.5 million podcasts globally. Although entries are labeled as Turkish or other languages, these tags were unreliable; therefore: 
despite the presence of language tags in the original 4.5M-podcast index, I first **concatenated title and description fields** and applied a **20-character minimum length filter** to the combined text, which **removed only ~20K entries** from the dataset.
I then ran **lingua-py** and **fastlangdetect** on the remaining data and retained only entries where **both detectors agreed on Turkish**, resulting in a **high-confidence, clean Turkish podcast subset**.

## What’s Inside

- **Source**: `podcastindex_feeds_lang_tr_both.parquet`
- **Rows**: 18,489
- **Format**: Parquet
- **Columns**: Full podcast feed metadata (RSS URL, title, description, language, enclosure URL, etc.) + language detection outputs.

## License

This dataset is released under the **Apache-2.0** license.

## Important Notice (Copyright)

- This dataset only contains **RSS feed links and metadata**.
- It does **not** include any podcast audio files.
- It is intended for research and indexing purposes.
- The dataset **does not distribute or re-host podcast content**, so it **does not violate podcast copyright**.
- If you do not want your podcast to be listed in this dataset and would like it to be removed, please contact us by opening a topic in the community section. Your request will be fulfilled.
## Example Usage

```python
import pyarrow.parquet as pq

table = pq.read_table("podcastindex_feeds_lang_tr_both.parquet")
print(table.schema)
```

## Citation

If you use this dataset, please cite the original source of the podcast feed index and this subset:

```bibtex
@dataset{solakoglu2026_turkish_podcast_rss,
  author = {Hasan Can Solakoğlu},
  title = {Turkish Podcast RSS Links},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/hcsolakoglu/turkish-podcast-links}
}