Tool

AI Explainability 360 Open Source Toolkit

A good way to get acquainted with the idea of explainable AI is to look at IBM Research's AI Explainability 360 Open Source Toolkit. In addition, they also offer an opportunity for consultation and a place to ask questions in the forum.

The toolkit offers various algorithms that you can use to make explainability part of your AI. According to IBM, the purpose of the toolkit is to ensure that what happens in the lab is also applied in other environments, such as health and education.

What you should know before reading further

  • Process Phase: Development
  • System component: Entire application and users
  • Price: Freely available

Method

There are several tutorials, demos and videos in this toolkit that clearly explain what explainable AI is about. An example of this is a demo that explains the different requirements of an explanation for a financial advisor and a customer. Through these kinds of exercises they try to translate lab concepts into situations that are recognizable.

In addition to information, this toolkit also has a number of algorithms that can be used to make an explanation part of your algorithm. The link on the website points to Github where the codes for these algorithms are available.

Result

This toolkit aims to make explainability a part of your AI. But if you do not yet know why or how this applies to your own AI, the toolkit offers an easy way to acquire this information.

Values as mentioned in the tool Related
ALTAI-principles
  • Explainability
  • Accountability
  • Fairness
  • Diversity, Non-discrimination & Fairness
  • Transparency

Link

AI Explainability 360 Open Source Toolkit

http://aix360.mybluemix.net/

This tool was not developed by the Knowledge Center Data & Society. We describe the tool on our website because it can help you deal with ethical, legal or social aspects of AI applications. The Knowledge Center is not responsible for the quality of the tool.