This is a generic library for building a simple web-based decision trees UI. Data is provided in JSON format, which is then displayed on a website. The tree is fully accessible both in terms of keyboard browsing and screen reading. It has been tested with all major screenreaders.

The code is in Typescript. After cloning the repository, just type

npm install
npm run-script build

This builds the decisiontree.js library in dist.

Alternative build is done using typescript and webpack:

npx tsc
npx webpack

Or simply install the package using

npm install html-decision-tree

Here is a very simple decision tree for deciding if you should play tennis:

Tennis example

Here is the corresponding flow chart decision tree for deciding if you should play tennis in SVG

and in JSON format.

Current node format in some pseudo grammar:

{
type: binary|nary|null,
value: 0-9+,
title: \w*,
content: \w*,
labels: [{
  text: \w*,
  value: 0-9+,
  }],
children: [node*],
action: URL
}

The idea is nary nodes have content (that is the description on the radio button) which is linked via the value to the children . Binary nodes have default description content as Yes (0) and No (1). Actions are only considered in leaf nodes and open a new Tab.

• Carlos Cavalie, ETS, for accessibility testing and design
• Brian Hochhalter for UI design and wireframe design and prototyping
• Danielle Vargas, W.W.Norton, for CSS styles, accessibility testing and design
• Amaya Webster, Benetech, for suggesting the initial idea at the 2020 DIAGRAM and ETS Code Sprint, providing an initial example and putting it into JSON.