Text-Color Hybrid Labeling for Multiclass Map Visualization: A Comparative Evaluation of Four Annotation Strategies

Research output: Chapter in Book or Report/Conference proceedingConference Proceedingpeer-review

Abstract

Prior work has identified the shortcomings of color-only encodings for maps with many categories, yet systematic comparisons of hybrid text-color strategies remain scarce. We therefore ran an 80-participant crowdsourced study on choropleth maps with 8-13 categories - approaching the 10-hue perceptual limit - to compare four annotation designs (Legend-Aside, Label-Fill, Label-Fit, Colored Label-Fill) across Count, Identify, Compare, and Rank tasks. Results show that the Label-Fit Map - with a single, large in-situ label - yields the highest accuracy and speed and ranks first in readability; Legend-Aside excels in simple counting and side-by-side comparisons. These findings deliver clear, task-specific guidelines for enhancing multiclass map readability and efficiency, informing the design of more effective map visualizations. All supplementary materials are available at our GitHub repository.

Original languageEnglish
Title of host publication18th International Symposium on Visual Information Communication and Interaction, VINCI 2025
EditorsGunter Wallner, James She, Michael Burch, Hai-Ning Liang
PublisherAssociation for Computing Machinery, Inc
ISBN (Electronic)9798400718458
DOIs
Publication statusPublished - 18 Dec 2025
Event18th International Symposium on Visual Information Communication and Interaction, VINCI 2025 - Linz, Austria
Duration: 1 Dec 20253 Dec 2025

Publication series

Name18th International Symposium on Visual Information Communication and Interaction, VINCI 2025

Conference

Conference18th International Symposium on Visual Information Communication and Interaction, VINCI 2025
Country/TerritoryAustria
CityLinz
Period1/12/253/12/25

Keywords

  • Crowd-source Experiment
  • Geographical Data
  • Map Visualization

Cite this