--- title: K-Recipe2Vec API emoji: 🍳 colorFrom: yellow colorTo: red sdk: docker pinned: false --- # 🍳 K-Recipe2Vec AI 기반 ν•œμ‹ μ‹μž¬λ£Œ λŒ€μ²΄ μΆ”μ²œ μ„œλΉ„μŠ€ ## πŸ“– Overview ν•œκ΅­ μš”λ¦¬μ—μ„œ μ‹μž¬λ£Œλ₯Ό λŒ€μ²΄ν•  수 μžˆλŠ” 재료λ₯Ό AIκ°€ μΆ”μ²œν•΄μ£ΌλŠ” μ›Ή μ„œλΉ„μŠ€μž…λ‹ˆλ‹€. Doc2Vecκ³Ό Word2Vec λͺ¨λΈμ„ ν™œμš©ν•˜μ—¬ μ‹μž¬λ£Œ κ°„μ˜ 의미적 μœ μ‚¬λ„λ₯Ό λΆ„μ„ν•©λ‹ˆλ‹€. ## πŸ”— Demo πŸš€ **[Live Demo](https://korea-recipe-ai.streamlit.app/)** - Streamlit Cloud 배포 ## ✨ Features - **πŸ₯¬ μ‹μž¬λ£Œ λŒ€μ²΄ μΆ”μ²œ**: μ—†λŠ” μž¬λ£Œμ— λŒ€ν•œ μœ μ‚¬ 재료 μΆ”μ²œ - **πŸ“Š 3D μ‹œκ°ν™”**: PCA 기반 재료 벑터 곡간 μ‹œκ°ν™” - **πŸ’° 가격 정보**: μž¬λ£Œλ³„ 가격 정보 제곡 - **☁️ μ›Œλ“œν΄λΌμš°λ“œ**: μΆ”μ²œ 재료 μ‹œκ°ν™” ## πŸ› οΈ Tech Stack | Category | Technologies | |----------|-------------| | **Frontend** | Streamlit | | **ML Models** | Gensim (Doc2Vec, Word2Vec) | | **Data Processing** | Pandas, NumPy | | **Visualization** | Plotly, Matplotlib, WordCloud | | **Database** | Supabase | | **Deployment** | Streamlit Cloud | ## πŸ“ Project Structure ``` k-recipe2vec/ β”œβ”€β”€ app.py # Main Streamlit application β”œβ”€β”€ logic.py # Core recommendation logic (if exists) β”œβ”€β”€ requirements.txt # Python dependencies β”œβ”€β”€ d2v.model # Doc2Vec trained model β”œβ”€β”€ w2v.model # Word2Vec trained model β”œβ”€β”€ price_rank.csv # Price data └── stats.pkl # Preprocessed statistics ``` ## πŸš€ Getting Started ### Prerequisites - Python 3.8+ - pip ### Installation ```bash # Clone the repository git clone https://github.com/nneans/k-recipe2vec.git cd k-recipe2vec # Install dependencies pip install -r requirements.txt # Run the app streamlit run src/app.py ``` ## πŸ“Š Model Information ### Doc2Vec Model - ν•œκ΅­ λ ˆμ‹œν”Ό 데이터 기반 ν•™μŠ΅ - λ ˆμ‹œν”Ό λ‹¨μœ„ λ¬Έμ„œ μž„λ² λ”© ### Word2Vec Model - μ‹μž¬λ£Œ κ°„ 의미적 μœ μ‚¬λ„ ν•™μŠ΅ - λŒ€μ²΄ κ°€λŠ₯ν•œ 재료 μΆ”μ²œμ— ν™œμš© ## 🀝 Contributing 버그 리포트, κΈ°λŠ₯ μ œμ•ˆ, PR ν™˜μ˜ν•©λ‹ˆλ‹€! ## πŸ“ License MIT License ## πŸ‘€ Author **Mingyun Kang** - GitHub: [@nneans](https://github.com/nneans)