The code should run without issues using Python versions 3.*.
Libraries neccessary to run the code:
- pandas
- sqlalchem
- numpy
- re
- nltk
- sklearn
- pickle
- To run ETL pipeline that cleans data and stores in database
python data/process_data.py data/disaster_messages.csv data/disaster_categories.csv data/DisasterResponse.db - To run ML pipeline that trains classifier and saves
python models/train_classifier.py data/DisasterResponse.db models/classifier.pkl - To run the web app.
python apps/run.py
You can then enter the flask app on http://0.0.0.0:3001/
Social Media is one important source for information about potential disasters. The huge amount of users leads to fast and detailed information as people tend to heavily use Social Media in case of major events. Unfortunately the result is a incredibly large number of messages which is hard to interpret and is impossible to scan by hand.
Therefore this projects categorizes given Twitter messages using machine learning in order to potentially support disaster responses in the future.
- data/ contains everything required for data preperation, including the data as .csv, a python script for the preperation and the final database
- models/ contains the the python script for model creation and the final classifier
- app/ contains files for the web data creation
Thanks to Figure Eight for their data on Twitter messages.