README.md

---
license: other
language:
- zh
size_categories:
- n<1K
task_categories:
- text-retrieval
- question-answering
- text-generation
- feature-extraction
task_ids:
- document-retrieval
- semantic-similarity-scoring
- open-domain-qa
- language-modeling
pretty_name: "Tibetan Muslims Knowledge Base & RAG Dataset"
tags:
- tibetan-muslims
- khache
- tibet
- chinese
- zh-Hans
- chinese-muslim
- islamic-culture
- muslim-human-rights
- public-web-archive
- rag
- document-retrieval
- semantic-search
- knowledge-base
- parquet
- dataset-viewer
- salaamalykum
configs:
- config_name: articles
  data_files:
  - split: train
    path: data/articles/train-*.parquet
  - split: validation
    path: data/articles/validation-*.parquet
  - split: test
    path: data/articles/test-*.parquet
  - split: subset
    path: data/articles/subset-*.parquet
- config_name: retrieval
  data_files:
  - split: train
    path: data/retrieval/train-*.parquet
  - split: validation
    path: data/retrieval/validation-*.parquet
  - split: test
    path: data/retrieval/test-*.parquet
  - split: subset
    path: data/retrieval/subset-*.parquet
- config_name: embeddings
  data_files:
  - split: subset
    path: data/embeddings/subset-*.parquet
---
# Tibetan Muslims Knowledge Base & RAG Dataset

Contains 22 public Simplified Chinese articles from the Salaam Alykum Tibetan Muslims topic, packaged as a production-ready Hugging Face dataset with Parquet splits, Markdown article files, retrieval rows, metadata indexes, and a lightweight embedding preview layer.

> [!TIP]
> **Human Readers / 普通读者：** Looking for normal article reading instead of raw data? Open **[Files and versions -> content](https://huggingface.co/datasets/qurancn/TibetanMuslims/tree/main/content)** and start with **[content/README.md](https://huggingface.co/datasets/qurancn/TibetanMuslims/blob/main/content/README.md)**. Example article: **[3348 克什米尔藏族穆斯林发生了什么？129个家庭的流亡与身份困境](https://huggingface.co/datasets/qurancn/TibetanMuslims/blob/main/content/3348-%E5%85%8B%E4%BB%80%E7%B1%B3%E5%B0%94%E8%97%8F%E6%97%8F%E7%A9%86%E6%96%AF%E6%9E%97%E5%8F%91%E7%94%9F%E4%BA%86%E4%BB%80%E4%B9%88129%E4%B8%AA%E5%AE%B6%E5%BA%AD%E7%9A%84%E6%B5%81%E4%BA%A1%E4%B8%8E%E8%BA%AB%E4%BB%BD%E5%9B%B0%E5%A2%83.md)**. For web-style rendering, use **[GitHub Pages](https://salaamalykum.github.io/TibetanMuslims/)**.

## Dataset Viewer

- Articles viewer: https://huggingface.co/datasets/qurancn/TibetanMuslims/viewer/articles/subset
- Retrieval viewer: https://huggingface.co/datasets/qurancn/TibetanMuslims/viewer/retrieval/subset
- Embedding preview: https://huggingface.co/datasets/qurancn/TibetanMuslims/viewer/embeddings/subset
- Embeddable viewer URL: `https://huggingface.co/datasets/qurancn/TibetanMuslims/embed/viewer/articles/subset`

Embed example:

```html
<iframe src="https://huggingface.co/datasets/qurancn/TibetanMuslims/embed/viewer/articles/subset" width="100%" height="560"></iframe>
```

The primary machine-readable files are Parquet shards under `data/`. The first columns are lightweight identifiers, titles, dates, previews, counts, and URLs; full `content_text`, `content_markdown`, and retrieval `text` are kept later in the schema to keep the Viewer preview usable.

## Splits and configs

| Config | Split | Rows | Purpose |
|---|---:|---:|---|
| articles | train | 18 | Article-level view for training/evaluation experiments with human review. |
| articles | validation | 2 | Article-level validation slice. |
| articles | test | 2 | Article-level held-out slice. |
| articles | subset | 10 | Lightweight preview subset for the Dataset Viewer and demos. |
| retrieval | train | 18 | RAG/search rows with one retrieval chunk per article. |
| retrieval | validation | 2 | Retrieval validation slice. |
| retrieval | test | 2 | Retrieval held-out slice. |
| retrieval | subset | 10 | Lightweight retrieval preview subset. |
| embeddings | subset | 22 | Public embedding preview layer using deterministic local character n-gram hashing. |

## Files

- `data/articles/*.parquet` — article-level rows with title, provenance, readable previews, full text, Markdown, hashes, and source URLs.
- `data/retrieval/*.parquet` — retrieval/RAG rows optimized for search, citation indexing, and QA experiments.
- `data/embeddings/subset-*.parquet` — lightweight embedding preview columns: `x`, `y`, `embedding_32`, and source metadata.
- `content/*.md` — human-readable Markdown article files.
- `metadata/article_index.csv` and `metadata/article_index.jsonl` — lightweight index for citation, SQL filtering, and external crawlers.
- `metadata/field_schema.json` — exact field names and Parquet types.
- `metadata/release-v1.0.0.json` — release manifest.
- `visualizations/embedding_preview.html` — static local scatter/table preview generated from the embedding subset.

## Field design for SQL/Data Studio

The dataset is shaped as a queryable product, not only a file dump. Common filter fields include `article_id`, `title`, `language`, `date_published`, `year`, `primary_topic`, `topic_tags_text`, `author`, `char_count`, `has_media`, `media_count`, `source_url`, `github_markdown_url`, and `content_sha256`. Use the Hugging Face Dataset Viewer, Data Studio, or SQL Console to filter by year, topic tag, author, media presence, or source URL.

## Format truth

This is an article corpus and RAG/retrieval corpus. It is **not** a native ShareGPT, Alpaca, chat, or instruction-response dataset. It may be converted into instruction-tuning examples later only if a separate conversion pipeline and new dataset card document that transformation truthfully.

## Source and provenance

Source topic: https://salaamalykum.com/cn/topic/%E8%97%8F%E6%97%8F%E7%A9%86%E6%96%AF%E6%9E%97

Each row includes source traceability through `source_url`, `canonical_url`, `date_published`, `date_modified`, `author`, `topic_tags`, and `content_sha256`. The GitHub public archive is https://github.com/salaamalykum/TibetanMuslims and the readable Pages site is https://salaamalykum.github.io/TibetanMuslims/.

## Cleaning and generation rules

- Records were extracted only from public Salaam Alykum topic/article pages.
- Article HTML was normalized into Markdown and plain text.
- Media URLs are kept as public links; media binaries are not republished in this dataset.
- Raw SQL backups, user tables, passwords, tokens, cookies, sessions, IP logs, private tickets, and server secrets are excluded.
- Content hashes are SHA-256 values for integrity and deduplication.
- Parquet shards were written with small row groups and page indexes for stable preview/search behavior.

## License and rights

`license: other`. Article bodies remain source-specific because many posts are translated or reposted public internet material. Do not assume commercial training rights for all article bodies. Generated metadata, schemas, indexes, and packaging files may be reused under the public repository notice where the maintainer has rights to license them.

## Recommended uses

- Public-interest search and citation indexing.
- RAG retrieval experiments with human review.
- Media monitoring and public-web archival research.
- Chinese-language Islamic, Tibetan Muslim/Khache, Tibet, Kashmir, Kalimpong, and Muslim human-rights topic exploration.

## Limitations and failure boundaries

- Small corpus: 22 articles and 22 retrieval chunks.
- Primary language: Simplified Chinese (`zh-Hans`).
- Some source posts are reposted or translated public internet material; always cite original source URLs where possible.
- This dataset is not a legal claim, not a private database, and not a comprehensive representation of all Tibetan Muslim-related material.
- Empty or very short source articles are preserved when they existed publicly, with hashes and metadata intact.
- The embedding preview uses deterministic local character n-gram hashing. It is useful for lightweight visualization and repeatable demos, but it is not a substitute for a production semantic embedding model.

## Versioning

Current release: `v1.0.0` for source batch `v2026.07.30-tibetan-muslims-topic-22`. Cite a version tag or commit revision for reproducible downstream work.

## Citation

```bibtex
@dataset{qurancn_tibetan_muslims_2026,
  author    = {Salaamalykum},
  title     = {Tibetan Muslims Knowledge Base & RAG Dataset},
  year      = {2026},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/datasets/qurancn/TibetanMuslims},
  version   = {v1.0.0}
}
```