Deployed via railway.app

https://turkish-nlp-app-production-451f.up.railway.app/

📋 Nasıl Çalışır (How It Works)

5-step workflow from text input to results Visual step-by-step process

🎯 Özellikler ve Teknikler (Features & Techniques) Detailed explanation of each feature with:

Tokenization - spaCy's Turkish tokenizer NER - Machine learning-based entity recognition POS Tagging - CRF-based tagger Dependency Parsing - Transition-based parser Lemmatization - Rule + dictionary-based Sentiment Analysis - TextBlob/Pattern library Word Cloud - WordCloud + Matplotlib Stop Word Detection - spaCy's Turkish stop words

💻 Teknolojiler (Technologies) Complete tech stack:

Backend: Python 3.10, Flask, Gunicorn, spaCy, TextBlob NLP Model: tr_core_news_md from Turkish NLP Suite

Training data: UD Turkish BOUN Treebank Model architecture: Transformer embeddings + CNN ~18,000 Turkish sentences

Frontend: Bootstrap 5.3, Bootstrap Icons DevOps: Docker, Railway, Git

🏗️ Sistem Mimarisi (System Architecture) ASCII diagram showing:

User → Gunicorn → Flask → NLP Layer → Response

⚡ Performans (Performance) Optimization details:

Model loading strategy Parallel processing Memory management UTF-8 encoding

📚 Akademik Kaynaklar (Academic References) Citations for:

spaCy Universal Dependencies Turkish BOUN Treebank TextBlob Turkish NLP Suite

The page is beautifully designed with:

🎨 Gradient backgrounds ✨ Hover effects on feature boxes 📱 Fully responsive 🔵 Bootstrap Icons 🌈 Color-coded sections