How mshknov/dipspam works
Spam Detector Pro offers multilingual single-message and CSV spam analysis through a Streamlit UI, plus a FastAPI prediction endpoint. Users provide text or a file, choose sensitivity in the UI or supply an API threshold, and receive a spam verdict; CSV users can download results. CatBoost is the classifier. The README documents capabilities not fully visible in the sampled UI excerpt; utility modules and configuration are shown separately where their wiring is evidenced. API prediction is directly evidenced; UI-to-model details beyond model loading are limited by the partial excerpt.
Main entry points: App.py Streamlit UI · api.py Prediction API · catboost_spam_model.cbm CatBoost Model · validators.py Input Validation
Components 12
User Interfaces
- Streamlit UI
App.py - Single Analysis
App.py - CSV Processing
App.py - Metrics View
App.py - API Integration View
App.py - Prediction API
api.py
Classification
- CatBoost Model
catboost_spam_model.cbm
Text Utilities
- Feature Extraction
utils/feature_extraction.py - Text Preprocessing
utils/preprocessing.py - Input Validation
utils/validators.py
Configuration
- Project Settings
config.py
Other components
- User
Connections 9
- User → Streamlit UI: uses
README.md - Streamlit UI → Single Analysis: offers
App.py - Streamlit UI → CSV Processing: offers
App.py - Streamlit UI → Metrics View: offers
App.py - Streamlit UI → API Integration View: offers
App.py - User → Prediction API: submits text
README.md - Prediction API → CatBoost Model: loads and predicts
api.py - Streamlit UI → CatBoost Model: loads model
- Input Validation → Project Settings: reads limits
utils/validators.py