InfoBayAI/Hindi_Non_STEM_QA_Dataset

Hindi Non-STEM Question Answering Dataset

Overview

The Hindi Non-STEM Question Answering Dataset is a comprehensive educational resource containing over 1.2 million Hindi question-answer pairs across humanities, commerce, economics, business studies, law, languages, social sciences, and general studies. The dataset is designed for Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), educational AI assistants, intelligent tutoring systems, document understanding, and multilingual NLP research.

Each question includes multiple-choice options, the correct answer, and a detailed explanation, making the dataset ideal for supervised fine-tuning, reasoning tasks, educational search, and AI-powered learning applications.

The dataset is available in both PDF and structured JSON formats.

---

Dataset Highlights

Property	Value
Language	Hindi
Total Questions	1.2M+
Question Type	Multiple Choice
Detailed Explanations	✅
File Formats	PDF & JSON
Average Question Length	~210 Words
AI Ready	✅

---

Subject Coverage

The dataset covers a broad range of Non-STEM educational disciplines.

Humanities

• History (इतिहास)
• Geography (भूगोल)
• Political Science (राजनीति विज्ञान)
• Civics (नागरिक शास्त्र)
• Sociology (समाजशास्त्र)
• Philosophy (दर्शनशास्त्र)
• Psychology (मनोविज्ञान)
• Social Science (सामाजिक विज्ञान)

Commerce & Business

• Commerce (वाणिज्य)
• Business Studies (व्यवसाय अध्ययन)
• Economics (अर्थशास्त्र)
• Management (प्रबंधन)
• Entrepreneurship (उद्यमिता)
• Marketing (विपणन)
• Accounting Principles (लेखांकन सिद्धांत)

Law & Governance

• Law (विधि)
• Indian Constitution (भारतीय संविधान)
• Public Administration (लोक प्रशासन)
• Governance (सुशासन)

Language & Communication

• General Hindi (सामान्य हिन्दी)
• Hindi Grammar (हिन्दी व्याकरण)
• Literature (साहित्य)
• Communication Skills (संचार कौशल)

General Studies

• General Knowledge (सामान्य ज्ञान)
• Current Affairs (समसामयिक घटनाएँ)
• Teaching (शिक्षण)
• Library & Information Science
• Home Science
• Environmental Studies
• Miscellaneous Educational Topics

---

Sample Record

{
  "question": "भारतीय संविधान का अनुच्छेद 14 किस अधिकार से संबंधित है?",
  "options": [
    "समानता का अधिकार",
    "स्वतंत्रता का अधिकार",
    "शिक्षा का अधिकार",
    "संवैधानिक उपचार का अधिकार"
  ],
  "answer": "समानता का अधिकार",
  "explanation": "अनुच्छेद 14 सभी नागरिकों को कानून के समक्ष समानता और कानून के समान संरक्षण का अधिकार प्रदान करता है।"
}

---

Dataset Features

• Over 1.2 million Hindi question-answer pairs
• Detailed answer explanations
• Multiple-choice questions
• Humanities and social science coverage
• Competitive examination content
• School and university curriculum
• AI-ready JSON annotations
• Human-readable PDF documents
• High-quality educational content
• Optimized for Hindi NLP applications

---

Data Structure

Hindi_NonSTEM_QA/
│
├── PDFs/
│   ├── History.pdf
│   ├── Geography.pdf
│   ├── Economics.pdf
│   ├── Law.pdf
│   └── ...
│
├── JSONs/
│   ├── History.json
│   ├── Geography.json
│   ├── Economics.json
│   ├── Law.json
│   └── ...

---

AI Tasks Supported

• Educational Question Answering
• Open-Domain Question Answering
• Retrieval-Augmented Generation (RAG)
• Supervised Fine-Tuning (SFT)
• Instruction Tuning
• Reading Comprehension
• Document Understanding
• Knowledge Retrieval
• Educational Search
• Hindi NLP
• Multilingual AI

---

Applications

This dataset is ideal for:

• Hindi Large Language Models (LLMs)
• AI Tutors
• Educational Chatbots
• Competitive Exam Preparation
• Government Exam Platforms
• Knowledge Retrieval Systems
• Digital Learning Platforms
• Enterprise Knowledge Bases
• Intelligent Search Systems
• Research in Hindi NLP

---

Industries

• Artificial Intelligence
• Education Technology (EdTech)
• Schools
• Universities
• Government Education
• Competitive Examination Platforms
• Digital Publishing
• Research Organizations
• Enterprise Learning
• Online Education

---

Advantages

• Large-scale Hindi educational corpus
• Rich explanatory content
• Covers major Non-STEM disciplines
• AI-ready structured format
• Optimized for Hindi language models
• Compatible with RAG pipelines
• Human-readable PDF references
• Structured JSON annotations
• Suitable for educational and reasoning-focused AI

---

Dataset Statistics

• Language: Hindi
• Questions: 1.2M+
• Subjects: 20+
• Question Format: Multiple Choice
• Explanations: Included
• Formats: PDF & JSON
• Average Question Length: ~210 words

---

Limitations

• Focused exclusively on Hindi-language educational content.
• Subject distribution varies across educational domains.
• Intended primarily for educational, research, and AI development purposes.

---

License

This dataset is distributed under the Creative Commons Attribution 4.0 International (CC BY 4.0) License.

Users are free to use, modify, distribute, and build upon the dataset with appropriate attribution.

---

Citation

@dataset{hindi_nonstem_qa_dataset,
  title={Hindi Non-STEM Question Answering Dataset},
  year={2026},
  license={CC BY 4.0}
}

---

Conclusion

The Hindi Non-STEM Question Answering Dataset is a large-scale educational corpus containing over 1.2 million Hindi question-answer pairs spanning history, geography, economics, commerce, business studies, law, political science, general knowledge, social sciences, and language learning. Featuring detailed explanations, structured JSON annotations, and PDF source documents, it is well suited for training Hindi Large Language Models, Retrieval-Augmented Generation systems, educational assistants, and multilingual AI applications.