Lan, Wu, Sun, Cheng, Shi, Zeng, Zhang and Pen – ASSESSING THE SUITABILITY OF FRACTAL DIMENSION FOR MEASURING GRAPHIC COMPLEXITY CHANGE IN SCHEMATIC METRO NETWORKS 2024

Description

This study tests whether fractal dimension (F), via box-counting analysis, quantifies complexity changes when metro maps are schematized. Twenty-six Chinese metro systems were analyzed in original and two schematic forms (Gaode multilinear and official octilinear). Fractal dimension decreased after schematization for every network (all log–log fits had adjusted R² > 0.998). Reductions in F correlated positively with original F (Spearman’s ρ ≈ 0.715 for Gaode, 0.853 for official), with bootstrapped confidence intervals confirming significance. A psychological questionnaire (40 respondents per method; 79 valid responses) showed perceived complexity change also correlated with measured reductions (ρ = 0.411 for Gaode, 0.687 for official). Comparison with Feature Congestion and Edge Density showed raw FC and ED can be misleading, but a locally averaged FC matched fractal results. The authors conclude fractal dimension is a robust metric for assessing schematization-induced complexity reduction and aligns with human judgment—especially for octilinear schematization—though no single schematization uniformly improves all networks.