How nadia-alsalem/financialnews works
Project type and purpose
• This is a full-stack machine-learning web application for Financial News Sentiment Classification.
• It provides a Streamlit front-end UI for user interaction and a FastAPI back-end (in src/FastAPI/api.py) that performs model inference using a pre-trained Logistic Regression model and TF-IDF vectorizer stored in save model/.
File structure
• Top-level files:
• Dockerfile, requirements.txt → containerization and dependencies
• data/news.csv → raw dataset
• save model/… → persisted model artifacts
• README.md → project overview
• src/ directory:
• src/data_processing.py/ → data cleaning and feature extraction logic
• src/data_analysis.py/ → EDA and reporting scripts
• src/model_training/ → model building and training workflows
• src/FastAPI/api.py → back-end service exposing a /predict endpoint
• No dedicated front-end folder in the tree, but the README describes a Streamlit app (presumably in a top-level or under src/).
README for additional insights
• The Streamlit UI: text input area, “Classify” button, visual effects (balloons or snow).
• Inference flow: input text → TF-IDF transform → Logistic Regression prediction → UI feedback.
• Key technologies: Python, Streamlit, FastAPI, scikit-learn, Docker.
Essential elements
• Component names (e.g., Streamlit App, FastAPI /predict).