Goal of this project is to design a web app that classifies a given twitter message into one or more of 36 cathegories.

The work consists of following steps:

1. ETL Pipeline: loads the training data, transforsms, cleans and saves it into an SQL database.
2. ML Pipeline: loads the data after ETL step, defines a machine learning pipeline, fits it to the data and save resulting model in pkl file
3. Web app: uses a ML Model created by ML Pipeline to predict a category of given tweet message.

In order to run this project following python libraries are needed: - ETL Pipeline: pandas, numpy, langid, sqlalchemy - ML Pipeline: pyndas, numpy, sqlite3, re, time, warnings, pickle, nltk, matplotlib, seaborn, sqlalchemy, sklearn, scipy - Web app: pandas, json, plotly, nltk, flask, sklearn, sqlalchemy

This project contains two jupyter notebooks: ETL_Pipeline_Preparation.ipynb and ML_Pipeline_Preparation.ipynb both were used to prepare ETL and ML pipelines described above. The actual ETL and ML pipelines are stored in data/process_data.py and models/train_classifier.py respectively. Source (raw) data that is used by ETL pipeline is stored in two csv files: data/disaster_categories.csv and data/disaster_messages.csv. The SQL databased used by ML pipeline is stored into a twitter_messages.db file. Finally, a web app can be found under app/run.py

1. Run the following commands in the project's root directory to set up your database and model.

   • 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

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

3. Go to localhost:3001

This project is a part of Udacity "Data Scientist" Nanodegree