/disaster-response

Project for Udacity's Data Scientist Nanodegree

Primary LanguageJupyter Notebook

Installations

In order to run the codes in this project, the following libraries must be installed:

  1. Pandas
  2. Numpy
  3. Sci-kit Learn
  4. Flask
  5. SQL Alchemy
  6. Plotly
  7. NLTK

Motivation

This project was done to complete the requirements for Udacity's Data Scientist Nanodegree. Using text data from Figure-8, a company specializing in data analytics and machine learning, the purpose was to classify messages that were created during a disaster into 36 categories to help in aid efforts.

Files

The project is divided into 3 folders: one for data and data processing; another one is for building a machine learning pipeline; and the third is for the web app. There are also 3 screenshots for the final web app.

Files in the Data Folder

  1. Messages data: disaster_messages.csv
  2. Categories data: disaster_categories.csv
  3. SQL Database: DisasterResponse.db
  4. Jupyter notebook for building ETL pipeline: ETL Pipeline Preparation.ipynb
  5. Python script for processing the data: process_data.py

Files in the Models Folder

  1. Jupyter notebook for building a machine learning pipeline: ML Pipeline Preparation.ipynb
  2. Python script for training the classifier: train_classifier.py
  3. A pickle file that contains the trained model: classifier.pkl

Files in the App Folder

  1. Python script for running the web app: run.py
  2. templates folder that contains 2 HTML files for the app front-end: go.html and master.html

Instructions for running the Python scripts:

  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 it: 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 http://0.0.0.0:3001/

Results

The final output of the project is an interactive web app that takes a message from the user as an input and then classifies it.

Screenshots

Web app Top 10 Categories Top 20 Words

Acknowledgement

Thanks to Udacity for providing guidance to complete the project and thanks to Figure-8 for providing the data