README.md

---
license: cc-by-sa-4.0
task_categories:
- text-retrieval
language:
- de
tags:
- wiktionary
- dictionary
- german
- linguistics
- morphology
- semantics
- normalized
- lossless
size_categories:
- 1M<n<10M
---

# German Wiktionary - FULL Normalized SQLite Database

This is a **complete, lossless, and fully normalized** SQLite database of German Wiktionary, capturing 100% of the structured data from the `cstr/de-wiktionary-extracted` dataset.

It is designed for production-ready applications, complex linguistic analysis, and mobile apps (Flutter, React Native) that require a comprehensive local dictionary.

## 🎯 Key Features

- **✅ 100% Lossless**: All 30+ top-level and nested fields from the source JSONL are preserved.
- **⚡ Fast Queries**: Fully indexed schema for sub-20ms queries.
- **🔗 Full Semantic Web**: Includes all semantic relations (synonyms, antonyms, **hypernyms, hyponyms, meronyms, holonyms, coordinate_terms**).
- **🗣️ Rich Content**: Includes **expressions, proverbs, and entry notes** in addition to definitions and examples.
- **📱 Mobile-ready**: Optimized for `sqflite` (Flutter) and other local DB use cases.
- **(and all features from the standard DB: forms, translations, sounds, etc.)**

## 📊 Database Statistics

- **Entries**: 970,801
- **Word Senses**: 3,098,364
- **Definitions (Glosses)**: 3,087,300
- **Translations**: 1,131,251
- **Word Forms (Inflections)**: 6,100,090
- **Form Tags (Total)**: 25,966,680
- **Pronunciations (Sounds)**: 2,327,762
- **Usage Examples**: 427,322
- **Synonyms**: 161,563
- **Antonyms**: 76,054
- **Hypernyms**: 133,059
- **Hyponyms**: 217,179
- **Proverbs**: 1,078
- **Expressions**: 13,138
- **Descendants**: 211
- **Entry Notes**: 16,536
- **Unique Tags**: 185
- **Unique Topics**: 58
- **Unique Categories**: 352

## 🏗️ Database Schema (Full)

This schema includes all tables from the standard `de-wiktionary-sqlite-normalized` dataset, plus the following additions:

- **entries**:
  - `title`: The Wiktionary page title.
  - `redirect`: The page this entry redirects to (if any).
- **entry_notes**: (New Table) Free-text notes associated with an entry (e.g., "Es gibt etliche Belege für die Steigerung...").
- **other_pos**: (New Table) Alternative part-of-speech values for this word.
- **entry_raw_tags**: (New Table) Unparsed, raw tags from Wiktionary.
- **descendants**: (New Table) Words in other languages descended from this word.
- **hypernyms**: (New Table) "Is-a" relationship (e.g., "Tier" is a hypernym of "Hund").
- **hyponyms**: (New Table) "Type-of" relationship (e.g., "Hund" is a hyponym of "Tier").
- **holonyms**: (New Table) "Part-of" relationship (e.g., "Hand" is a holonym of "Finger").
- **meronyms**: (New Table) "Has-a" relationship (e.g., "Finger" is a meronym of "Hand").
- **coordinate_terms**: (New Table) Sibling terms (e.g., "Hund" and "Katze" are coordinate terms under "Haustier").
- **expressions**: (New Table) Idiomatic expressions using the word (linked to `sense_id`).
- **proverbs**: (New Table) Proverbs using the word (linked to `sense_id`).

*(For the standard schema, see the `cstr/de-wiktionary-sqlite-normalized` dataset card)*

## 📖 Usage

### Download
```python
from huggingface_hub import hf_hub_download
import sqlite3
import gzip
import shutil

# Download compressed database
db_gz_path = hf_hub_download(
    repo_id="cstr/de-wiktionary-sqlite-full",
    filename="de_wiktionary_normalized_full.db",
    repo_type="dataset"
)

# Decompress (if it's .gz)
db_path = db_gz_path.replace('.gz', '')
with gzip.open(db_gz_path, 'rb') as f_in:
    with open(db_path, 'wb') as f_out:
        shutil.copyfileobj(f_in, f_out)

# Connect
conn = sqlite3.connect(db_path)
```

### Example Query (New Tables)
```python
# Get all hypernyms (parent categories) for "Hund"
cursor.execute('''
    SELECT h.hypernym_word
    FROM entries e
    JOIN hypernyms h ON e.id = h.entry_id
    WHERE e.word = ? AND e.lang = 'Deutsch'
''', ('Hund',))

print("Hypernyms of 'Hund':", [row[0] for row in cursor.fetchall()])
```

## 🔗 Source

Original data: [cstr/de-wiktionary-extracted](https://huggingface.co/datasets/cstr/de-wiktionary-extracted)

## 📜 License

CC-BY-SA 4.0 (same as source)