TajikNLPWorld/tajik-morphological-rules

Исправь таблиуц ## 📊 Dataset Statistics он не видно :

---

language: - tg license: apache-2.0 tags: - morphological-rules - prefixes - suffixes - tajik-language - nlp - stemming - lemmatization datasets: - tajik-morphological-rules

🇹🇯 Tajik Morphological Rules (Prefixes & Suffixes)

A complete, production-ready set of prefix and suffix rules for Tajik morphological analysis. These rules power lemmatization, stemming, and morphological segmentation.

📖 Description

This dataset provides two configuration files:

• prefix_rules.json — 100+ prefix stripping rules with priorities and safety flags.
• suffix_rules.json — 150+ suffix stripping rules, ordered from longest chains to shortest.

Each rule is annotated with:

• form: the affix to strip
• replacement: what to replace it with (usually empty)
• safe: whether it can be applied without POS/dictionary checks
• requires_dictionary: whether a dictionary lookup is needed
• priority: processing order (higher = earlier)
• group: logical category (e.g., "CORE_VERBAL_GRAMMATICAL")

📊 Dataset Statistics

Metric	Count
Total prefix rules	58
Safe prefix rules	48
Total suffix rules	54
Safe suffix rules	40
Total rules	112

📁 Data Format

The dataset is organized as a Hugging Face DatasetDict with two configurations:

Config	Description	Rows
prefix_rules	All prefix stripping rules	58
suffix_rules	All suffix stripping rules	54

Each row contains:

Field	Type	Description
form	string	The affix to match
replacement	string	Replacement string (usually empty)
kind	string	"prefix" or "suffix"
group	string	Rule group (e.g., "1_CORE_VERBAL_GRAMMATICAL")
safe	bool	Whether rule is safe without context
requires_dictionary	bool	Whether dictionary validation is needed
strip	bool	Whether to strip in standard lemmatization
priority	int	Processing order (higher = first)

🚀 Usage

Load with 🤗 Datasets

from datasets import load_dataset

# Load prefix rules
prefixes = load_dataset("TajikNLPWorld/tajik-morphological-rules", "prefix_rules")
print(prefixes["train"][0])

# Load suffix rules
suffixes = load_dataset("TajikNLPWorld/tajik-morphological-rules", "suffix_rules")
print(suffixes["train"][0])

Use in a stemmer

def apply_rules(word, rules):
    for rule in rules:
        if word.startswith(rule["form"]):   # for prefixes
            return word[len(rule["form"]):]
    return word

📜 License

Apache License 2.0

🤝 Citation

@dataset{tajik_morphological_rules,
    title = {Tajik Morphological Rules (Prefixes & Suffixes)},
    author = {TajikNLPWorld},
    year = {2025},
    publisher = {Hugging Face},
    url = {https://huggingface.co/datasets/TajikNLPWorld/tajik-morphological-rules}
}

---

Generated with ❤️ for the Tajik NLP community