Definition
A Treemap is a data visualization tool that allows you to view hierarchical data represented as colored rectangles of varying sizes. Treemaps display these rectangles, which represent different categories in a hierarchical structure, in a shape similar to that of a tree. Viewing data in this way makes it possible for you to see and compare patterns and inconsistencies [1]. Let’s take a look at an example of a Tree Map with the figure below.
Tree map of an AI
training data distributionThe above is a treemap
visualization that shows how an AI system's training data is distributed across
different categories. The largest rectangle represents text data (45% of the
total dataset), which is further subdivided into web pages (25%), books (20%),
social media (15%), news articles (25%), and scientific papers (15%). The next
largest category is image data (25%), split between photos and diagrams.
Smaller rectangles represent audio data (10%), video data (8%), and specialized
data types (12%).
Origin
The origins of
treemap goes as far back as the early 1990s, when Ben Shneiderman, a computer
scientist, introduced this innovative visualization process as a way to display
hierarchical data structures with efficiency and clarity. They became well
received due to the fact that they make it easy to comprehend complex data.
They have changed and now have cool designs and features to make them even
better. They began as a tool to organize files, and now people use them a lot
in finance, market research, and other areas. People who work with data find
treemaps helpful [2].
Treemaps are
used by professionals in industries such as health care, finance, business,
politics, and research to visualize and understand complex sales trends and
patterns and efficiently represent medical data [1].
Why it Matters
Treemaps are
important in AI and data visualization because they effectively display
hierarchical data in a compact, space-efficient way, allowing users to quickly
identify patterns, proportions, and outliers within complex datasets. Treemaps
are also very versatile. You can use them across multiple industries to quickly
create an effective visualization of large amounts of complex data, which may
not be possible with other charts and diagrams. If your data is hierarchical,
with distinct numerical values, and you’re looking to display only positive
values, you can create a treemap.
Related Terms
- Hierarchical Data: Treemaps are designed to visualize data that has a hierarchical structure, meaning data organized in levels or categories.
- Nested Rectangles: The core visual element of a treemap is the use of rectangles, with larger rectangles representing higher-level categories and smaller rectangles representing subcategories, creating a nested or "tree-like" structure.
- Size and Color Encoding: The size of each rectangle often represents the magnitude or proportion of a data value, while color can be used to differentiate categories or highlight specific attributes.
In Practice
A real-life case
study of a company practicing the use of tree maps in AI can be seen in the case of Amazon with
their product recommendation engine. Amazon uses tree-based algorithms and tree
map visualizations in their recommendation systems. Amazon's use of tree-based
models showcases how companies can leverage hierarchical data structures to
organize complex product relationships and customer behavior data to drive
sales through targeted recommendations.
References
- Coursera Staff. (2025). What Is a Treemap?
- Pyne, C. (n.d). Whatis a Treemap and How to Create it?