From Top-Right to User-Right: Perceptual Prioritization of Point-Feature Label Positions

TL;DR

This paper addresses the lack of consensus on Point-Feature Label Positioning PPOs by introducing PerceptPPO, a user-validated, perceptually grounded PPO derived from a global study with nearly 800 participants. It employs a rigorous 2AFC experimental design and Thurstone-based modeling to rank eight candidate label positions around anchors, revealing a strong user preference for labels positioned above features and challenging traditional top-right emphasis. The work also investigates optimal label density, finding a mean preference around 12.5% LD_thr and 16% local density, and demonstrates that PerceptPPO outperforms traditional PPOs in perceived quality, with Zoraster performing comparably in some analyses. The findings offer practical guidance for cartographic and GIS applications, supporting a shift toward user-centered label placement that can enhance readability and user engagement in visualizations.

Abstract

In cartography, Geographic Information Systems (GIS), and visualization, the position of a label relative to its point feature is crucial for readability and user experience. Alongside other factors, the point-feature label placement (PFLP) is typically governed by the Position Priority Order (PPO), a systematic raking of potential label positions around a point feature according to predetermined priorities. While there is a broad consensus on factors such as avoiding label conflicts and ensuring clear label-to-feature associations, there is no agreement on PPO. Most PFLP techniques rely on traditional PPOs grounded in typographic and cartographic conventions established decades ago, which may no longer meet today's user expectations. In contrast, commercial products like Google Maps and Mapbox use non-traditional PPOs for unreported reasons. Our extensive user study introduces the Perceptual Position Priority Order (PerceptPPO), a user-validated PPO that significantly departs from traditional conventions. A key finding is that labels placed above point features are significantly preferred by users, contrary to the conventional top-right position. We also conducted a supplementary study on the preferred label density, an area scarcely explored in prior research. Finally, we performed a comparative user study assessing the perceived quality of PerceptPPO over existing PPOs, advocating its adoption in cartographic and GIS applications, as well as in other types of visualizations. Our research, supported by nearly 800 participants from 48 countries and over 45,500 pairwise comparisons, offers practical guidance for designers and application developers aiming to optimize user engagement and comprehension, paving the way for more intuitive and accessible visualizations.

Figures (10)

• Figure 1: Common label positions for a point feature as presented in prior literature: \ref{['fig:fixed-4-model']} 4-position model, \ref{['fig:fixed-8-model']} additional positions of 8-position model, and \ref{['fig:fixed-10-model']} additional positions of 10-position model. Abbreviations: T -- top, B -- bottom, R -- right, L -- left, S -- slightly. Compounds, as TSR -- top slightly right, consists of the former trivial position names.
• Figure 2: Example of map area rendered consistently using TR and T position across all point features.
• Figure 3: Results of the PerceptPPO user study. Chart \ref{['fig:ppo-quality-score']} depicts quality z-scores and 95% confidence intervals. Chart \ref{['fig:ppo-significance']} shows a triangle plot visualizing significant differences between label position preferences as proposed by Pérez-Ortiz and Mantiuk Perez2017. Each red circle represents a label position, and the lines indicate significant differences between pairs. Solid blue lines represent statistically significant differences. The edge values show the absolute difference in z-scores between the compared label positions, with the $p$-values denoted in brackets. The label positions are plotted along the x-axis, with alternating y-axis offsets for clarity.
• Figure 4: Identified clusters of users in PerceptPPO study with depicted quality z-scores and 95% confidence intervals. The red dashed brackets, with the corresponding $p$-values reported above, represent the PPOs pairs without evidence of statistically significant differences at $\alpha=0.05$ using the two-tail test. Conversely, between the PPOs pairs without brackets are detected statistically significant differences.
• Figure 5: Renders of map area 0 using PercepPPO at various values of $LD_\mathrm{thr}$ employed in the evaluation of label density described in \ref{['sec:label-density']}.