Understanding Bias in Perceiving Dimensionality Reduction Projections

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Doh, Seoyoung, Jeon, Hyeon, Shin, Sungbok, Quadri, Ghulam Jilani, Kim, Nam Wook, Seo, Jinwook
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912505436045312
author Doh, Seoyoung
Jeon, Hyeon
Shin, Sungbok
Quadri, Ghulam Jilani
Kim, Nam Wook
Seo, Jinwook
author_facet Doh, Seoyoung
Jeon, Hyeon
Shin, Sungbok
Quadri, Ghulam Jilani
Kim, Nam Wook
Seo, Jinwook
contents Selecting the dimensionality reduction technique that faithfully represents the structure is essential for reliable visual communication and analytics. In reality, however, practitioners favor projections for other attractions, such as aesthetics and visual saliency, over the projection's structural faithfulness, a bias we define as visual interestingness. In this research, we conduct a user study that (1) verifies the existence of such bias and (2) explains why the bias exists. Our study suggests that visual interestingness biases practitioners' preferences when selecting projections for analysis, and this bias intensifies with color-encoded labels and shorter exposure time. Based on our findings, we discuss strategies to mitigate bias in perceiving and interpreting DR projections.
format Preprint
id arxiv_https___arxiv_org_abs_2507_20805
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Understanding Bias in Perceiving Dimensionality Reduction Projections
Doh, Seoyoung
Jeon, Hyeon
Shin, Sungbok
Quadri, Ghulam Jilani
Kim, Nam Wook
Seo, Jinwook
Human-Computer Interaction
Machine Learning
Selecting the dimensionality reduction technique that faithfully represents the structure is essential for reliable visual communication and analytics. In reality, however, practitioners favor projections for other attractions, such as aesthetics and visual saliency, over the projection's structural faithfulness, a bias we define as visual interestingness. In this research, we conduct a user study that (1) verifies the existence of such bias and (2) explains why the bias exists. Our study suggests that visual interestingness biases practitioners' preferences when selecting projections for analysis, and this bias intensifies with color-encoded labels and shorter exposure time. Based on our findings, we discuss strategies to mitigate bias in perceiving and interpreting DR projections.
title Understanding Bias in Perceiving Dimensionality Reduction Projections
topic Human-Computer Interaction
Machine Learning
url https://arxiv.org/abs/2507.20805