The primary aim of this research is to design, develop, and evaluate a framework for real-time sentiment analysis of social media data to support and enhance crisis management decision-making in Nigeria.
lagos_flood_sentiment/
│
├── app.py # Streamlit UI (dashboard only)
├── models.py # Loading & prediction logic (SVM + BERT)
├── data_pipeline.py # Dataset loading, synthetic generation, future real-tweet ingestion
├── config.py # Paths, keywords, locations, constants
├── requirements.txt
├── svm_pipeline.pkl
├── bert_lagos_model/ # HF model + tokenizer
├── synthetic_lagos_floods.csv
└── .streamlit/
└── config.toml # theme config