alexanderhipp/disaster-response-pipeline-udacity-project

Project Submission for Udacity Nanodegree

Disaster Response Pipeline Project

The project is part of the Data Scientist Udacity nanodegree. Thanks to figure-eight for providing the nessecary data which consists of real messages that have been sent during a specific disaster. The Pipeline is an NLP machine learning algorithm that takes and processes these messages and categorizes them by content. A web application is also being used to conveniently show with category the message is.

What packages I have used:

This project was developed using:

• flask 1.0.2
• joblib 0.13.1
• nltk 3.4
• pandas 0.24.1
• plotly 3.5.0
• scikit-learn 0.20.2
• sqlalchemy 1.2.17

How to run the project:

1. To set up your database and model, run following commands in the project's root directory.

   • To run the ETL pipeline that cleans data and stores in database python data/process_data.py data/disaster_messages.csv data/disaster_categories.csv data/disaster_response.db
   • To run the ML pipeline that trains classifier and saves python models/train_classifier.py data/disaster_response.db models/disaster_response

2. Run following command in the app's directory to run the web app. python run.py.

   Please note that for run.py to run properly the database needs to be called disaster_response.db and the model disaster_response.joblib.

3. Go to http://0.0.0.0:3001/