Contributing

From The Foundation for Best Practices in Machine Learning

Welcome to the contribution guide for the FBPML Best Practices, thanks for checking out this page, we are looking forward to your contributions!

What can I contribute?

You can contribute to the Best Practices in the following ways:

  • edit existing items (controls and aims) - Everyone is allowed to edit the item pages. Any changes will be 'pending' until review, but everyone will still see unreviewed changes. The items on the overview pages (the ones with tables) are not editable, because they don't actually contain any text: they are filled from the item pages automatically .
  • propose new items (controls and aims) - Anyone can post in our Forum to propose a new item. The reason we ask you to do it like this, is because the overview pages are 'magic' pages and only editable by administrators.
  • add elaborations, examples or your own experience - Anyone can add any additional information to any item page under the 'Additional Information' header. These changes will be 'pending' until review, but everyone can still see your changes. As long as your additional information does not violate any of our policies, we will be quick to accept your additions. You could also add your own experience about your journey into implementing the Best Practices under the User Guides page.
  • make and discuss suggestions about the structure of the Best Practice, in other words suggestions impacting multiple items or sections, on our Forum
  • anything else - If you have a good idea that we didn't cover above, feel free to contact us or to post in the Forum!

Before suggesting additions (a new control, or an addition to an existing control), please try to check if your addition has not already been included in a different section. The FBPML Best Practices are already pretty large, so we try not to repeat any controls in multiple places. The rule of thumb we employ to determine which section a control belongs to is that it goes in the section where the control is the most general.

How does it work?

  1. (optional) Create a Wiki account. This is highly recommended but not necessary. This way you only have to solve a Captcha during sign-up, and not for every edit you do.
  2. Find the item you want to edit, go to its own page. For example, Data Subjectivity (which is located under Technical Best Practices > Representativeness & Specification, neither of which are directly editable themselves).
  3. Click "Edit" on the bar on the top right.
  4. Change what you want to change. Note that changes to the Control and Aim sections will be reviewed more carefully than changes to the Additional Information. In the bar at the top of the editor window you can find editing help, like how to create lists, links and such.
  5. Write a little summary of your change, then click "Save changes" at the bottom of the page. Your changes are now 'Pending' until a reviewer approves the changes, but everyone will still be able to see your changes.
  6. An added benefit is that it's very easy to see the difference between the versions of every page under "View History" or the "Pending Changes"

Policies

Please take note of our Code of Conduct and follow them in all edits, posts and similar.

License

Our Best Practices are open-source and community driven, and have therefore been licensed under the Creative Commons Attribution license. This means they are freely available for commercial and/or private adaptation and/or use, subject to attributing The Foundation of Best Practices of Machine Learning.

Note also that if you do contribute to our Best Practices, you accede that your work will be transferred to The Foundation for Best Practices in Machine Learning so that we can, subsequently, license it under the Create Common Attribution license in order to make it freely available to all. Should you have any more questions in this regard, please feel free to contact us.