🇰🇭 Khmer Lexicon Dataset

Dataset Summary

A comprehensive Khmer language lexicon dataset containing 12,653 unique terms with rich linguistic metadata, semantic relationships, and domain classifications. This specialized collection focuses on formal Khmer vocabulary including royal terminology, government positions, Buddhist religious terms, and technical vocabulary.

🇰🇭 100% Khmer Definitions: All definitions are written in Khmer script, providing native language understanding for Khmer NLP models. No English translations - pure Khmer-to-Khmer semantics.

📊 Key Statistics

Total Words: 12,653
Semantic Relationships: 30,242 relationships
POS Distribution: 98.1% NOUN (specialized lexicon characteristic)
Domains: Royal, governmental, religious, technical, and more
Language: Khmer (km)

🎯 Intended Uses

This dataset is designed for:

Tokenizer Training: Building Khmer-specific tokenizers with cultural awareness
NLP Model Development: Training language models for Khmer
Linguistic Research: Studying Khmer morphology and semantics
Dictionary Systems: Building comprehensive Khmer dictionaries
Cultural Preservation: Preserving formal and royal Khmer vocabulary

📁 Dataset Structure

Data Files

The dataset includes multiple Parquet files for different use cases:

Main Tables:
- words.parquet: Complete word list with all metadata
- synsets.parquet: Semantic synsets with definitions
- relations.parquet: Semantic relationships between synsets
- lexicon_combined.parquet: Combined view for easy analysis
Training Splits:
- train.parquet: 80% of data (10,122 samples)
- validation.parquet: 10% of data (1,265 samples)
- test.parquet: 10% of data (1,266 samples)

Data Fields

Each entry contains:

word: The Khmer term
pos: Part of speech tag
frequency: Frequency score
weight: Importance weight (0-100)
domain: Domain classification
register: Language register (formal, colloquial, royal, etc.)
syllable_count: Number of syllables in the word
char_length: Number of characters in the word
definition: Contextual definition in Khmer language (e.g., "នាមបញ្ជាក់វត្ថុឬគំនិត")
semantic_field: Semantic category
outgoing_relations: Number of outgoing semantic relationships
incoming_relations: Number of incoming semantic relationships

🏷️ POS Tags

The dataset uses the following POS tags:

NOUN: Nouns (12,418 terms, 98.1% - specialized lexicon characteristic)
VERB: Verbs (84 verified terms, 0.7%)
ADJ: Adjectives (40 terms, 0.3%)
PRON: Pronouns (26 terms, 0.2%)
ADV: Adverbs (24 terms, 0.2%)
CLAS: Classifiers (22 terms, 0.2%)
NUM: Numbers (22 terms, 0.2%)
PART: Particles (17 terms, 0.1%)

Total non-NOUN terms: 235 (all manually verified)

🌐 Domains

Major domains include:

Royal and palace terminology
Government and administrative terms
Buddhist and religious vocabulary
Legal and judicial terms
Military and police terminology
Medical and health vocabulary
Technology and modern terms
Core vocabulary and common words

📝 Important Notes

Why 98% NOUN?

This is a specialized lexicon, not a general corpus. It contains:

Royal titles and palace vocabulary
Government positions and institutions
Proper names and entities
Technical and domain-specific terminology

These categories are naturally noun-heavy. All 235 non-NOUN terms have been carefully verified and represent genuine verbs, adjectives, pronouns, etc.

🔗 Semantic Relationships

The dataset includes 30,242 semantic relationships:

Antonyms: 25,992 pairs
Co-occurrence: 2,000 patterns
Morphological: 500 relations
Synonyms: 500 pairs
Classifier-for: 400 relationships
Domain-related: 400 connections
Hypernym: 200 (is-a relationships)
Meronym: 150 (part-of relationships)
Numerical order: 100 sequences

💾 File Sizes

Total dataset size: ~15 MB
Words table: 5.3 MB
Synsets table: 2.8 MB
Combined lexicon: 6.6 MB
Relations table: 14 KB

📜 License

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

🙏 Acknowledgments

This lexicon was developed for advancing Khmer NLP and preserving the rich vocabulary of the Khmer language, with special attention to formal, royal, and cultural terminology.

📚 Citation

If you use this dataset, please cite:

@dataset{khmer_lexicon_2025,
  title={Khmer Lexicon: A Comprehensive Linguistic Dataset},
  author={Khmer NLP Community},
  year={2025},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/[username]/khmer-lexicon}
}

🤝 Contributing

To contribute to this dataset or report issues, please visit the repository or contact the maintainers.

Dataset Version: 1.0.0
Last Updated: 2025-08-19

Datasets:

khopilot
/

khmer-lexicon

You need to agree to share your contact information to access this dataset