🇰🇿 Question-Answering_Kazakh

A comprehensive Kazakh-language question-answer dataset for fine-tuning and training language models.
Created and maintained by Kurumikz. Free to use with attribution.

---

📌 Overview

Question-Answering_Kazakh is an open-domain QA dataset written entirely in the Kazakh language (kk). It covers a wide range of topics — from the history and geography of Kazakhstan to Kazakh grammar, culture, economy, and language learning (Kazakh ↔ English).

The dataset was built with fine-tuning large language models in mind. Each record consists of a natural-language question and a detailed, informative answer. Answers range from short factual statements to comprehensive multi-paragraph explanations — giving the model exposure to a variety of response lengths and styles.

🔗 Part of the Zerde Collection

This dataset is a focused subset of the larger Zerde-QA-50K dataset — a synthetically generated corpus of 51,422 Kazakh QA pairs spanning 20+ academic and professional domains.

The current dataset (1,424 records) concentrates specifically on Kazakhstan-centric topics: history, geography, culture, language, economy, science, and sports — making it a compact, high-precision resource for domain-specific fine-tuning on Kazakh national knowledge.

If you need broader topical coverage (IT, AI/ML, cybersecurity, law, medicine, philosophy, and more), consider using the full Zerde-QA-50K →.

---

📊 Dataset Statistics

Metric	Value
Total records	1,424
File format	JSONL
File size	~1.9 MB
Primary language	Kazakh (kk)
Duplicate records	0
JSON parse errors	0
Records with null/empty fields	0

Answer Length Distribution

Length range	Records	Share
< 100 chars	39	2.7%
100 – 400 chars	948	66.6%
400 – 1000 chars	177	12.4%
1000+ chars	260	18.3%

Field Length Stats (in characters)

Field	Min	Max	Avg	Median
question	15	263	67	59
answer	32	7,111	692	292

---

🗂️ Format

The dataset is stored as JSONL (JSON Lines) — one JSON object per line.
Every record has exactly two fields:

{
  "question": "Қазақстанның астанасы қай қала?",
  "answer": "Қазақстан Республикасының астанасы – Астана қаласы. Бұрын бұл қала Нұр-Сұлтан, одан бұрын Ақмола, Целиноград деп аталған. Астана – еліміздің саяси, әкімшілік және іскерлік орталығы."
}

Field	Type	Description
question	string	Natural-language question in Kazakh
answer	string	Detailed answer in Kazakh

No null values. No empty strings. No duplicate records.
Encoding: UTF-8. No BOM.

---

📚 Content Areas

The dataset covers 8 major thematic domains:

🏛️ 1. History of Kazakhstan

• Kazakh Khanate — founding, rulers, decline
• Golden Horde and its legacy
• Key historical figures: Abylai Khan, Kenesary, Abylay
• Russian Empire colonization (1731–1916)
• The 1916 national uprising
• Alash movement and Alashorda government
• Soviet era: collectivization, famine (1930–1933)
• December 1986 uprising (Jeltoqsan)
• Path to independence (1991)

🌍 2. Geography

• Major cities: Almaty, Astana, Shymkent, Karaganda
• Rivers: Ili, Irtish, Syr Darya, Ural
• Lakes and seas: Caspian Sea, Aral Sea, Balkhash
• Mountains: Tian Shan, Altai, Khan Tengri
• Steppes and natural zones: Saryarka, Betpak-Dala
• National parks: Ile-Alatau, Karaganda, Katon-Karagay

🎭 3. Culture & Traditions

• National holidays: Nauryz, Independence Day
• Traditional customs: wedding ceremonies, betashar, zheti ata
• National cuisine: beshbarmak, kymyz, kurt, baursak
• Music and instruments: dombra, kuy, zhyrau tradition
• Epic poetry: Koblandy Batyr, Er Targyn, Alpamys
• Poets and writers: Abai, Mukhtar Auezov, Zhambyl

📖 4. Kazakh Language & Linguistics

• Alphabet history: Arabic script → Latin → Cyrillic → Latin again
• Grammar: agglutination, vowel harmony (syngarmonia)
• All 7 grammatical cases with English comparisons
• Suffix system in depth:
  • Plural suffix -lar/-ler
  • Possessive suffixes
  • Personal/conjugation suffixes
  • Case suffixes
  • Word-forming suffixes: -shы/-shi, -lyq/-lik, -syz/-siz, -day/-dey, -la/-le, -ғan/-gen, -atyn/-etin
• Tense system vs English 12 tenses
• Negation, imperatives, conditionals
• SOV word order vs English SVO
• Postpositions vs English prepositions
• Kazakh ↔ English translation pairs and sentence examples

💰 5. Economy & Industry

• Oil & gas: Tengiz, Kashagan, Karachaganak fields
• Mining & metallurgy: Karaganda steel, Balkhash copper
• Uranium: Kazatomprom — world's largest producer
• Agriculture: virgin lands campaign, grain exports
• EXPO 2017 "Future Energy" in Astana
• Silk Road history and modern Belt & Road Initiative

🏛️ 6. Politics & Governance

• Constitution of Kazakhstan (1995, 2022 reforms)
• Parliament: Senate and Mazhilis
• Presidential system and powers
• SCO (Shanghai Cooperation Organisation) — Kazakhstan's role
• Nuclear disarmament — Budapest Memorandum 1994
• Astana Process — Syria peace negotiations
• Jeltoqsan 1986 — political significance

🔬 7. Science & Technology

• Baikonur Cosmodrome — history and significance
• Sputnik launch (1957), Gagarin flight (1961)
• Kazakh cosmonaut Aidyn Aimbetov
• Semey nuclear test site — ecological consequences
• Nazarbayev University
• Kazakhstan's nuclear-free status

🏅 8. Sports & Notable Figures

• Gennady Golovkin (GGG) — world boxing champion
• Serik Sapiyev — Olympic gold, Val Barker Trophy
• Kazakh traditional sports: kokpar, baige, kazaksha kures

---

💡 Intended Use

This dataset is designed for:

• Fine-tuning causal language models (GPT-style) on Kazakh-language QA
• Instruction tuning — teaching models to follow question→answer format
• Retrieval-Augmented Generation (RAG) — as a knowledge base
• Evaluation of Kazakh NLP models
• Educational tools — Kazakh language learning applications
• Benchmarking — testing models' knowledge of Kazakhstan

For a significantly larger and more topically diverse dataset from the same author, see Zerde-QA-50K (51,422 records, 20+ domains).

Recommended training format

For instruction fine-tuning, wrap records like this:

### Сұрақ:
{question}

### Жауап:
{answer}

Or in English-style instruct format:

<|user|>
{question}
<|assistant|>
{answer}

---

⚙️ Loading the Dataset

With 🤗 Hugging Face datasets

from datasets import load_dataset

ds = load_dataset("kurumikz/Question-Answering_Kazakh")
print(ds["train"][0])
# {'question': '...', 'answer': '...'}

Manually with Python

import json

records = []
with open("data.jsonl", "r", encoding="utf-8") as f:
    for line in f:
        records.append(json.loads(line))

print(f"Loaded {len(records)} records")
print(records[0])

With pandas

import pandas as pd

df = pd.read_json("data.jsonl", lines=True)
print(df.shape)          # (1424, 2)
print(df.columns)        # ['question', 'answer']
print(df['answer'].str.len().describe())

---

🔍 Data Quality

The dataset was programmatically verified before release:

Check	Result
JSON parse errors	✅ 0
Empty lines	✅ 0
Missing question or answer	✅ 0
Null / None values	✅ 0
Exact duplicate records	✅ 0
Duplicate questions	✅ 0
Duplicate answers	✅ 0
Control characters	✅ 0
BOM markers	✅ 0
Encoding errors (UTF-8)	✅ 0
Whitespace-only values	✅ 0

Character script breakdown (total 1,081,140 chars):

Script	Count	Share
Cyrillic	848,017	78.4%
Latin	20,409	1.9%
Digits	7,194	0.7%
Arabic	2	~0.0%

The 2 Arabic characters appear on one line intentionally —
used as phonetic examples in a Kazakh linguistics answer.

---

📜 License

This dataset is released under the
Creative Commons Attribution 4.0 International (CC BY 4.0) license.

You are free to:

• ✅ Use — for any purpose, including commercial
• ✅ Share — copy and redistribute
• ✅ Adapt — remix, transform, build upon

Under the condition:

• 📌 Attribution required — You must give appropriate credit
  to the original author when using or distributing this dataset.

Suggested citation:

Question-Answering_Kazakh dataset, available at:
https://huggingface.co/datasets/kurumikz/Question-Answering_Kazakh
Licensed under CC BY 4.0.

---

🙏 Attribution

If you use this dataset in research, a product, or any public project,
please include a visible credit such as:

"This work uses the Question-Answering_Kazakh dataset,
created by Kurumikz,
available at huggingface.co/datasets/kurumikz/Question-Answering_Kazakh"

For questions, feedback, or collaboration — reach out:
kurumikz — kurumikaz@gmail.com

---

🗃️ Related Datasets

Dataset	Records	Scope	Link
Question-Answering_Kazakh (this)	1,424	Kazakhstan-focused: history, geography, culture, language	—
Zerde-QA-50K	51,422	20+ broad domains: IT, AI, law, medicine, humanities & more	→ View on HF

---

Made with ❤️ for the Kazakh NLP community.