Data Viz / UXUI

Donut Charts Exploration

Complex security data often requires multiple levels of context. This project explored donut and nested donut visualizations as a way to represent bot traffic distributions, relationships, and hierarchy within analytics dashboards. The work focused on component design, interaction behavior, and visual clarity to help users interpret large datasets more effectively.

Client F5
Role UX Design / Data Visualization / Component Design
Publication 2025 Bot Defense / F5 Distributed Cloud Services

Challenge

Problem

Complex security datasets often require multiple levels of context, but traditional donut charts flatten hierarchical relationships into a single view. This made it difficult to explore parent-child relationships without sacrificing readability or overwhelming users with information.

Solution

A progressive disclosure model was explored using nested donut charts, expandable hierarchy, and contextual legends. The component reveals additional levels of detail only when needed, allowing users to move from overview to investigation while maintaining context.

Exploration

Overview

Multiple concepts explored hierarchy, legend organization, and interaction patterns for complex, nested data.

Hierarchy

Explored approaches for visualizing nested data while preserving context. Concepts focused on progressive disclosure, clear parent-child relationships, and maintaining orientation as users navigated deeper into the hierarchy.

A. Overview

Top-level categories.

B. First Expansion

Reveal second level.

C. Deep Hierarchy

Reveal third level.

D. Collapse State

Return one level.

E. Reset

Return to overview.

Legends

Explored legend structures for organizing nested data while balancing readability, hierarchy, and scanability.

A. Stacked Legend

  • Separate levels

  • Clear organization

  • Repeated labels

B. Tree Indentation

  • Parent-child grouping

  • Reduced repetition

  • Strong hierarchy

C. Simplified Indentation

  • Cleaner layout

  • Less visual noise

  • Faster scanning

Interaction Patterns

Explored interaction models for revealing nested data while balancing discoverability, flexibility, and interface simplicity.

A. Interactive Legend

  • Multiple entry points

  • Flexible exploration

  • Higher interaction density

B. Non-Interactive Legend

  • Single entry point

  • Simpler behavior

  • Cleaner interface

Outcome

The exploration established a reusable component system for visualizing hierarchical data. The final concept balances readability, scalability, and progressive disclosure while preserving context across multiple levels.

Previous
Previous

Motion / Event
Women Leaders in AI

Next
Next

Motion / Video
AI for Customer Service