Skip to main content

Taxonomy Principles

To help ensure that the open taxonomy will enable data scientists to collect better data and model more effectively, we’ve created a set of core principles. Any new development or update will be checked against these principles.

The open taxonomy should:

  1. Give data scientists the best possible data to work with for modeling.
  2. Be supported by tooling to implement it, with as little manual work as possible.
  3. Be 100% covered by automated validation.
  4. Work for any website, web-app or app built on modern, well adopted frameworks.
  5. Be consistent across events, contexts, frameworks, and platforms.
  6. Have a clear intended instrumentation & modeling use for each element.
  7. Never create unnecessary complexity or inheritance.

These principles are a starting point and we’ll adapt them along the way if that proves to be necessary to reach the taxonomy goals. They are open to feedback and contribution.