This repository contains resources that can be used to integrate ethics into a data science project. Specifically, the repository contains the following files:
.
├── checklist.md # An AI ethics checklist
├── dataset_card_template.md # A template for documenting datasets
├── model_card_template.md # A template for documenting machine learning models
├── resources.md # A file with resources that were used to create the checklist
└── README.md
The biggest part of the checklist is based on Deon. I adapted the original checklist and added additional points, based on the resources you can find here.
Using a checklist is a first step towards making machine learning projects more ethical. However, documenting the model itself and the dataset that was used to train and evaluate the model is really important, too. Therefore I provide a template for a model card and a template for a dataset card.
A ton of resources on the topic of AI Ethics exist. The sources I used to create the checklist contained in this repo can be found here.