twitter-disaster-classification

An application that scrapes tweets off of Twitter based on given tags and uses a trained classification model to predict disasters.

How to get started

• Get yourself a Twitter Developer API key by signing up here.
• Copy the contents from the sample.env file into a .env file and fill its contents.
• The AUTH_TOKEN can be anything. It is used for validation purposes.

Dependencies

• Download all the dependencies by running $ pip install -r requirements.txt

Training

• You can train the model by running $ python model_setup.py
• Once this is done, you will be able to see a disaster_model.sav file.

Server

• You can run the server by running $ python main.py

Client

• Run the client manually and interact with the application there.

Tests

• You can run the tests by running $ python test_main.py

Demos

demo.mov

Further Scope

• First and foremost, the model needs to be trained on a better and/or larger dataset.
• The UI needs to be improved.
• Add data visualization in the client side.
• Heatmaps (need to get location information).
• Get more useful information out of the scraping process (will work on this).
• Please do open issues if you can think of any!

Bugs?

• Feel free to open an issue.