Discrimination Web App - Detecting discrimination through Suppes-Bayes Causal Network

A bacherlor thesis by Blai Ras. Read the full paper at https://drive.google.com/file/d/12ta5mAYn_H_jlzm7jM6DsEvHPYLxJsX6/view?usp=sharing

Running the Project Locally

Essential requirements

First, install git:

	sudo apt install git

Second, check your current Python version with:

	python3 -V

This project needs Python >= 3.5

Third, install pip3:

	sudo apt install python3-pip

Project specific requirments

Install R on your computer. Check if your already have it:

	R --version

This project needs R >= 3.2.3

If not, first add the project GNU Privacy Guard key

	sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys E298A3A825C0D65DFD57CBB651716619E084DAB9

Add the R Repository:

	sudo add-apt-repository 'deb https://cloud.r-project.org/bin/linux/ubuntu bionic-cran35/'

Update & install:

	sudo apt update
	sudo apt install r-base

On the first run project will tell you if you want to install the required bnlearn R library. You can type 'yes' or install it manually on R before launching the project, with:

		install.packages(“bnlearn”)

---

Clone the repository to your local machine:

	git clone https://github.com/rohit-nlp/discrimination-web-app.git

---

Install the requirements:

	pip3 install -r requirements.txt

Some of the requirements cannot be installed through pip.

Pycairo

	pip3 install pycairo
	pip3 install cairocffi

or

	sudo apt-get install python3-cairocffi

iGraph

	pip3 install python-igraph

If (1) doesn't work try:

	sudo apt install libxml2-dev libz-dev

And try (1) again.

---

Apply the migrations:

	python3 manage.py migrate

Finally, run the development server:

	python3 manage.py runserver

The project will be available at http://127.0.0.1:8000/