• The Human Settlement Quality (HSQ) scores of thousands of communities were evaluated. • Significant differences were observed in residents’ perceptual distances to various factors. • The model with optimized perceptual scales effectively enhances assessment accuracy. • Chinese residents prioritize socio-economic factors, while Germans focus on natural factors. • The livability deficiencies of each community were revealed through HSQ shortcoming maps. Human Settlement Quality (HSQ) is a critical component of sustainable urban development, directly affecting residents’ health, well-being, and quality of life. However, most existing studies rely on expert-defined service radii and indicator weights at the city scale, overlooking residents’ perceptions and failing to capture fine-grained variations at the community level. This study proposes a Perceptual-Scale Optimized Random Forest (PSO-RF) to enable human-centered, community-scale HSQ evaluation by integrating subjective satisfaction data with objective environmental indicators. The framework captures multi-scale perceptual differences in environmental features and determines the optimal measurement scale for each indicator, leading to more accurate HSQ assessments. Five case cities in China and Germany—Beijing, Changsha, Shenzhen, Berlin, and Munich—were selected to reflect diverse regional, socio-economic, and developmental contexts, based on multi-source spatiotemporal data from 2010, 2015, and 2020. The findings reveal that: (1) Residents perceive HSQ across two dominant spatial scales: local (860 m) and accessible (2050 m); (2) Chinese communities emphasize socio-economic conditions within close proximity, while German communities prioritize broader natural environmental factors; (3) The PSO-RF model reduces evaluation error by 9.6% compared to fixed-radius approaches by identifying indicator-specific perceptual scales; (4) The generated HSQ and shortcoming maps uncover localized human settlement challenges and offer practical guidance for targeted urban planning. This study advances the methodological foundation for perception-driven livability research and provides actionable insights for precision urban governance.
lei et al. (Sun,) studied this question.