---
license: other
license_name: commercial-paid
language:
- lzh
- zh
pretty_name: Classical Chinese Variant Collation (校勘) — Commercial
task_categories:
- text-generation
- text-classification
tags:
- classical-chinese
- 校勘
- textual-criticism
- variant-collation
- 文献学
- commercial-dataset
size_categories:
- n<1K
configs:
- config_name: sample
  data_files: sample.jsonl
---

# Classical Chinese Variant Collation · 校勘  💰 (Commercial Dataset)

> **This is a commercial dataset.** A free **50-work preview sample** is provided
> below (`sample.jsonl`, texts truncated); the **full set with complete aligned
> texts is available upon request**.
>
> 📧 **To license / purchase or request a quote, email `wangeksy@gmail.com`.**
> ✅ Cleared for **commercial use** — both members are public-domain classical works.

## What this is
A textual-criticism dataset: classical works that survive in **two independent
editions** — a digital-canon recension and its **四库全书 (Siku Quanshu) imperial
edition** (plus some cross-canon pairs) — aligned as `(copy A, copy B)` pairs with a
character-level overlap score. These capture genuine **transmission variants**:
different recensions, abridgements, 繁简/异体字 differences, and editorial lineages of
the same work — not identical duplicates, not title-only collisions.

```json
{"task":"variant_collation","title":"明儒学案","overlap":0.83,
 "a_canon":"史藏","b_canon":"四库全书","a_hanzi":786545,"b_hanzi":770000,
 "a_text":"…full text of recension A…","b_text":"…full text of recension B…"}
```

## What you get (full set)
- **653 aligned variant pairs** (650 distinct works), each with the
  **complete text of both editions** (the public sample truncates to a preview).
- **629 pairs** align a treasury recension against its 四库全书 imperial
  edition; the remainder are cross-canon pairs within the treasury.
- Overlap band **0.45–0.985** (median **0.827**) — true variants only.
- Per-pair metadata: `title`, `overlap`, `a_canon`/`b_canon`, `a_dynasty`/`b_dynasty`,
  `a_hanzi`/`b_hanzi`. Parquet (zstd). Full list in `CONTENTS.md`.

## Why it is unique
- **No public corpus ships these pre-aligned.** Both members are PD base text; the
  product is the *alignment* and the variant signal.
- Spans canonical works: 元史, 唐会要, 明儒学案, 周礼注疏, 续名医类案, 元诗选, 宋诗钞, …
- Useful for: textual-criticism / 文献学 modeling, robustness to recensional &
  繁简/异体字 variation, deduplication research, 校异 study.

## This repo (listing)
Public **50-work preview sample** only (`sample.jsonl`, texts clipped). Full set with
complete aligned texts delivered on purchase — **email `wangeksy@gmail.com`**.

## Related
Companion commercial datasets from the same corpus: **文言文断句·标点 Punctuation
Restoration** (5.3M pairs) and an expert-graded **文言→白话 Translation + QA eval**.