⚠️ Private Repository - This is a private project. Access is limited to authorized contributors only.

A sophisticated web application for discovering Armenian rhymes based on IPA (International Phonetic Alphabet) phoneme similarity analysis. The app uses intelligent feature-aware algorithms to identify different types of rhymes.

• Multiple Rhyme Categories:

  • Perfect Rhymes: Identical phonetic tails
  • Near-Perfect Rhymes: High similarity (≥85%) with same vowel nucleus
  • Assonance/Near Rhymes: Same vowel nucleus with decent similarity (≥60%)
  • Slant Rhymes: Consonant-driven similarity (≥50%)

• Smart Phoneme Analysis:

  • Extracts rhyme tails (nucleus + coda of the last syllable)
  • Feature-aware Levenshtein distance for phoneme similarity
  • Considers consonant place/manner and vowel height/backness

• Modern Web Interface:

  • Beautiful, responsive design
  • Real-time autocomplete search
  • Detailed word definitions and part-of-speech tags
  • Similarity scores for each match
  • Mobile-friendly layout

• Python 3.7+
• Modern web browser (Chrome, Firefox, Safari, Edge)

pip install -r requirements.txt

Or manually:

pip install Flask==2.3.0 Flask-CORS==4.0.0 Werkzeug==2.3.0

Ensure dictionary-hy-improved.jsonl is in the project directory. Each line should be a JSON object with:

{
  "": "word",
  "p": ["part_of_speech"],
  "d": ["definition"],
  "f": ["word_form_1", "word_form_2", ...],
  "f_ipa": ["phonetic_form_1", "phonetic_form_2", ...]
}

python backend.py

You should see:

Loaded [count] words from dictionary
 * Running on http://127.0.0.1:5000

Open your browser and navigate to:

file:///c:/Users/Nazani/Desktop/armenian-rhyme-app/index.html

Or serve it with a local HTTP server:

# Using Python's built-in server
python -m http.server 8000

Then visit: http://localhost:8000

1. Rhyme Tail Extraction:

   • Identifies the last vowel in a word's phonetic representation
   • Extracts all phones from that vowel to the end (nucleus + coda)

2. Similarity Scoring:

   • Uses feature-aware Levenshtein distance
   • Identical phones: cost 0
   • Similar phones (same place/manner for consonants, similar height/backness for vowels): cost < 1
   • Unrelated phones: cost 1
   • Insertions/deletions: cost 1

3. Classification:

   • Perfect: Identical tails
   • Near-perfect: ≥85% similarity + same nucleus
   • Assonance: Same nucleus + ≥60% similarity
   • Slant: ≥50% similarity
   • None: Everything else (filtered out)

Consonants are classified by:

• Place: labial, alveolar, dental, palatal, velar, glottal
• Manner: stop, fricative, nasal, approximant

Vowels are classified by:

• Height: close, close-mid, open-mid, open, near-open
• Backness: front, central, back

Find rhymes for a word.

Parameters:

• word (required): Armenian word to find rhymes for
• limit (optional): Maximum results (default: 50)

Response:

{
  "word": "գլուխ",
  "rhymes": [
    {
      "word": "լուծ",
      "similarity": 0.857,
      "label": "near-perfect",
      "definition": "solution",
      "part_of_speech": "noun"
    }
  ]
}

Search for words by prefix.

Parameters:

• q (required): Search query

Response:

{
  "results": [
    {
      "word": "word",
      "definition": "meaning",
      "pos": "noun"
    }
  ]
}

Get detailed info about a word.

Response:

{
  "word": "բառ",
  "definition": "word",
  "pos": "noun",
  "forms_count": 25
}

Get dictionary statistics.

Response:

{
  "total_words": 18831
}

1. Click in the search field
2. Type an Armenian word (autocomplete suggestions will appear)
3. Press Enter or click "Search"
4. Results are sorted by rhyme quality (perfect → near-perfect → assonance → slant)

import requests

response = requests.get(
    'http://localhost:5000/api/rhymes',
    params={'word': 'գլուխ', 'limit': 25}
)
rhymes = response.json()
for rhyme in rhymes['rhymes']:
    print(f"{rhyme['word']} ({rhyme['label']}): {rhyme['similarity']}")

armenian-rhyme-app/
├── backend.py                       # Flask backend with rhyme algorithm
├── index.html                       # Web interface (frontend)
├── requirements.txt                 # Python dependencies
├── README.md                        # This file
├── dictionary-hy-improved.jsonl      # Armenian word dictionary
└── add_ipa.py                       # (Existing) IPA phonetization script

• Dictionary loading: ~2-3 seconds (18,000+ words)
• Per-search latency: 50-200ms depending on result count
• Autocomplete search: 100-300ms
• Browser queries are cached by the frontend

• Multi-language support (transliteration)
• Advanced filters (filter by part of speech, definition)
• Export results (CSV, JSON)
• Offline mode with service workers
• Word frequency ranking
• Rhyme suggestions for poetry composition
• Audio pronunciation playback

"Word not found" error:

• Check spelling and diacritics
• Use autocomplete to verify correct spelling

CORS errors:

• Ensure backend is running on port 5000
• Check that Flask-CORS is installed

Slow performance:

• First search loads dictionary (~3 seconds)
• Subsequent searches are much faster
• Try limiting results to 25-50

Port already in use:

# Find and kill process on port 5000
netstat -ano | findstr :5000
taskkill /PID <PID> /F

• Framework: Flask 2.3.0
• CORS: Flask-CORS 4.0.0
• Algorithm: Feature-aware Levenshtein distance
• Data Format: JSONL (JSON Lines)

• HTML5 with semantic markup
• CSS3 with gradients and animations
• Vanilla JavaScript (ES6+)
• No external dependencies (except for API calls)

• Dictionary loading: O(n) where n = number of words
• Per-search: O(n × m²) where m = average rhyme tail length
• Typical: ~0.1-0.2 seconds for 18,000 words

This project is provided as-is for educational and personal use.

Feel free to:

• Report bugs and issues
• Suggest improvements
• Add new rhyme categories
• Enhance the phoneme feature system

Check the API documentation above or examine the code comments in backend.py for implementation details.