myX-Burmese-Morpho-Synthetic

myX-Burmese-Morpho-Synthetic is a high-volume, synthetically augmented dataset consisting of over 37.8 million rows of Burmese word formations. Developed by Khant Sint Heinn (Kalix Louis) under the DatarrX organization, this resource is designed to advance the structural understanding of the Burmese language in the field of Natural Language Processing (NLP).

📌 Purpose

The primary goal of this dataset is to improve Burmese NLP by providing a massive scale of grammatically structured word combinations. While many of the generated strings may be semantically "meaningless" (synthetic), they strictly follow the morphological rules of the Burmese language. This makes it an invaluable resource for tasks that require structural analysis rather than semantic understanding.

👥 Target Audience

• NLP Researchers: For training and benchmarking Burmese-specific models.
• Computational Linguists: To study Burmese affixation and word construction.
• AI Developers: For building tools such as Stemmers, Lemmatizers, and Morphological Analyzers.

🛠 How it was Created (Methodology)

This dataset was generated using a Rule-Based Synthetic Augmentation approach. We utilized high-quality root words from open-source dictionaries and combined them with various linguistic markers:

1. Permutation Logic: We applied Cartesian products between Root words (Verbs, Nouns, Adjectives, Adverbs, and Names) and their respective Suffixes/Particles.
2. Morphological Patterns:
   • Verb Conjugation: Root Verbs + Tense/Aspect/Mood markers.
   • Noun Attributions: Nouns + Postpositional markers.
   • Negation: Systematic application of the "မ...ဘူး" (Negation) framework. Note: Some complex negations may contain minor edge-case variations.
   • Name Entity Linkage: Burmese Names + Functional particles to simulate real-world usage in sentences.

📊 Dataset Statistics

• Total Rows: 37,888,919
• Format: JSONL
• Columns:
  • root: The base word (Lemma).
  • suffix: The affix or functional particle attached.
  • combined: The final generated Burmese word/phrase.

📜 Technical Note on Semantics

This dataset is Synthetically Structured. This means that while the word formations are grammatically plausible according to Burmese morphology, the resulting "combined" strings may not carry real-world meaning. It focuses on the form of the language rather than the content.

⚖️ Usage & License

This dataset is released under the Apache License 2.0.

Key Terms:

• Freedom of Use: You are free to use, modify, and distribute this dataset for both academic and commercial purposes.
• Redistribution: If you redistribute this dataset or any derivative works, you must include a copy of the Apache 2.0 License.
• No Warranty: This dataset is provided "as is," without warranties or conditions of any kind.
• Contribution: We encourage the Burmese language community to utilize this resource to build better NLP tools and research.

For the full license text, please refer to Apache License 2.0.

🤝 Credits & Acknowledgments

We would like to thank the following creators for providing the foundational data used in this augmentation:

• Myanmar Words (8 Parts of Speech): myanmartools/myanmar-words
• Burmese Names Dataset: Hein Htet Ahkar

📝 Citation

If you use this tokenizer in your research or project, please cite it as follows:

APA 7th Edition

Khant Sint Heinn. (2026). *myX-Burmese-Morpho-Synthetic: A 37M-Row Morphologically Augmented Dataset (Version 1.0) * [Dataset]. Hugging Face. https://huggingface.co/datasets/DatarrX/myX-Burmese-Morpho-Synthetic

BibTeX

@dataset{datarrx2026morpho,
  author = {{Khant Sint Heinn}},
  title = {myX-Burmese-Morpho-Synthetic: A 37M-Row Morphologically Augmented Dataset},
  year = {2026},
  version = {1.0},
  publisher = {DatarrX},
  organization = {DatarrX},
  url = {https://huggingface.co/datasets/DatarrX/myX-Burmese-Morpho-Synthetic},
  note = {Released under DatarrX Open Source Initiative}
}

Acknowledgement & Credits

This dataset is developed by Khant Sint Heinn (Kalix Louis) and released officially as part of the DatarrX Open Source Initiative. While the individual developer holds the primary credit for the methodology, the resource is maintained and distributed by DatarrX.

Created by: Khant Sint Heinn (Kalix Louis) Organization: DatarrX