2026-08-28

7 arXiv categories· 96 journals· 2371 candidates — 9 worth your time· 76 without an open abstract

Headline

Immigrant exposure and right-wing voting dynamics across the Netherlands

Today, together

tag shift
no tag ran above its 30-day average
shares references
no items share references today counted over the 9 papers published above
on the same shoulders
5 of the day's papers cite Travel and the Built Environment (Reid Ewing, Robert Cervero 2010); 4 cite Travel demand and the 3Ds: Density, diversity, and design (Robert Cervero, Kara M. Kockelman 1997) counted over the 78 of this day's 85 named papers whose reference lists we have — standing on the same older work is not the same as being about the same thing
first appearance
Flinders University; Ho Chi Minh City University of Transport; Observatoire de Physique du Globe de Clermont-Ferrand; and 3 more none of them on a paper we published in the 77 days before this issue

Two counts, not one counted twice: every row here but “on the same shoulders” counts the papers published above, and that one counts every paper this day named — those and the ones under Also published today, whose abstracts we could not read.

datapreprint

Benedikt Meylahn, Tommaso Giommoni, Mike Lees, et al.
arXiv physics.soc-ph

Electoral Systems and Political Participation · Populism, Right-Wing Movements · Migration, Refugees, and Integration · Netherlands

The complex relationship between anti-immigrant sentiment and exposure in the Netherlands

This study analyzed the relationship between immigrant exposure and anti-immigrant sentiment across the Netherlands using national grid data and voting results from the 2021 and 2023 national elections. Using a novel exposure indicator adapted from residential segregation metrics, the analysis found that while municipalities and neighborhoods with higher overall immigrant exposure had lower right-wing vote shares in 2023, areas experiencing rapid increases in immigrant exposure between 2021 and 2023 showed a reversal, with rising right-wing vote shares. The research also identified that wealth and inequality levels influence these voting dynamics at high exposure thresholds.

Why it matters — It demonstrates that the spatial relationship between immigrant presence and political backlash is dynamic rather than static, revealing an 'overexposure effect' where rapid demographic change triggers right-wing voting shifts even if highly diverse areas remain generally less right-wing.

Caveat: The study relies on right-wing voting outcomes as a proxy for anti-immigrant sentiment, which may capture other political or economic grievances.

arXiv · pdf

published: Discover Cities

Fatemeh Aghnaei Alanjagh, Raziyeh Teimouri
Discover Cities

Tabriz Islamic Arts University

Urban Green Space and Health · Urban Heat Island Mitigation · Land Use and Ecosystem Services · machine learning · deep learning · Tabriz Islamic Arts University

Exploring the impact of AI on public open and green space planning to optimize urban sustainability

This study synthesizes literature and case studies to develop a conceptual framework for integrating artificial intelligence into the planning, design, and management of public open and green spaces. It evaluates how machine learning, deep learning, GeoAI, and neural network models like CNNs and LSTMs can be combined with GIS and multi-criteria decision-making to optimize green space allocation and predict environmental performance.

Why it matters — It establishes a structured pathway for urban planners to transition from static green space mapping to dynamic, predictive modeling, helping to identify underserved areas and adapt infrastructure to climate and social demands.

Caveat: The study is a conceptual synthesis and framework development rather than an empirical application or testing of a specific AI model in a real-world city.

Discover Cities · doi

published: Case Studies on Transport Policy

Hiep Tuan Nguyen, Lam Canh Nguyen, Huy Gia Dinh, et al.
Case Studies on Transport Policy

Ho Chi Minh City University of Transport

Maritime Ports and Logistics · Maritime Transport Emissions and Efficiency · Vehicle Routing Optimization Methods · Ho Chi Minh City University of Transport · RMIT Vietnam · Observatoire de Physique du Globe de Clermont-Ferrand

Mapping the impact of Industry 4.0 technologies on port sustainability: A systematic literature review

This study conducts a systematic literature review of 51 peer-reviewed journal articles indexed in Scopus up to July 2025 to analyze the role of Industry 4.0 technologies in port sustainability. Using bibliometric mapping and thematic synthesis, it categorizes dominant technologies like the Internet of Things, artificial intelligence, machine learning, and automation, and proposes a three-layer framework linking these technologies to economic, environmental, and social outcomes.

Why it matters — It establishes that current smart port research is heavily skewed toward operational efficiency, energy management, and emissions reduction, leaving a critical gap in evidence regarding social sustainability, workforce adaptation, and stakeholder inclusion.

Caveat: The findings are based exclusively on peer-reviewed journal articles indexed in Scopus, omitting insights from conference proceedings, technical reports, policy papers, and industry-related studies.

Case Studies on Transport Policy · doi

published: Habitat International

Gabriele D’Adda, Elisabetta Vitale Brovarone
Habitat International

Politecnico di Torino

Urban Planning and Governance · Urban Transport and Accessibility · Urban, Neighborhood, and Segregation Studies · Barcelona · Politecnico di Torino

Spatial and social justice on and beyond the streets: reflections on Barcelona's mobility and housing policies

This study analyzes Barcelona's municipal policies between 2015 and 2023, examining how the local government attempted to coordinate sustainable mobility initiatives, specifically the superblock experiments, with housing policies designed to curb gentrification and increase public housing supply.

Why it matters — It documents a concrete policy attempt to resolve the tension between urban greening and housing displacement, showing how mobility and housing strategies can be intentionally integrated to protect vulnerable residents from the inflationary effects of public space improvements.

Caveat: The analysis is a qualitative case study of a single city under a specific municipal administration, which may limit its direct application to cities with different governance structures or housing market dynamics.

Habitat International · doi

preprint

Haiyan Hao
arXiv cs.CY

Urban Transport and Accessibility · 3D Modeling in Geospatial Applications · Urban Design and Spatial Analysis · large language model · street network data · points of interest

Spatial-Knowledge-Graph-Grounded LLM Agents for Neighborhood Livability Evaluation

The researchers developed a prototype framework that combines a spatial knowledge graph with large language models to simulate and evaluate neighborhood livability through synthetic household schedules. Tested in a neighborhood in Shenzhen, the system integrates residents, facilities, and road networks to generate daily schedules, perform rule-based feasibility checks, and calculate actual travel paths and times. The resulting simulated events were used to conduct synthetic resident-agent interviews regarding daily convenience, travel burdens, and household coordination.

Why it matters — The framework demonstrates that nominal facility proximity does not guarantee actual accessibility, revealing that residents with limited mobility and care responsibilities experience disproportionate travel burdens. This provides a new, auditable method to evaluate urban livability by simulating how diverse household constraints and schedules interact with the physical built environment, moving beyond static spatial indicators.

Caveat: The framework's utility is demonstrated through a prototype simulation in a single neighborhood in Shenzhen, and the simulated experiences remain distinct from observed, real-world human perceptions.

arXiv · pdf

preprint

Pedro Lemes Sixel Lobo, Eric Tokuda, Kuruvilla Joseph Abraham, et al.
arXiv physics.soc-ph

Food Security and Health in Diverse Populations · Obesity, Physical Activity, Diet · Health disparities and outcomes · random forest · gradient boosting · São Paulo

District-Level Food Environment Indicators and Social Vulnerability in São Paulo

This study analyzed the relationship between food retail availability and social vulnerability across 76 districts in So Paulo, Brazil, using data from the Relao Anual de Informaes Sociais and CAISAN. Eight machine learning classifiers were trained to predict binary social vulnerability levels based on the densities of healthy and unhealthy food outlets and street markets, with XGBoost achieving the highest mean F-score of 0.75. In a Random Forest model, the combined densities of healthy and unhealthy food establishments accounted for roughly 60% of the feature importance.

Why it matters — It demonstrates that publicly available administrative data on food retail can successfully map onto district-level social vulnerability, offering a potential proxy for identifying socioeconomically disadvantaged areas when direct census data is unavailable.

Caveat: The findings are constrained by a small ecological sample size, class imbalance, and a cross-sectional design that prevents causal or household-level interpretations.

arXiv · pdf

preprint

Man Luo, Yixuan Zhao
arXiv cs.LG

Urban Transport and Accessibility · Human Mobility and Location-Based Analysis · Transportation and Mobility Innovations · graph neural network · New York City

Toward Equitable Low-Carbon Mobility: Fairness-Aware Demand Prediction for Expanding Bike-Sharing Systems

The proposed FairGIN framework is a fairness-aware graph neural network designed to predict bike-sharing demand for newly added stations. It simulates network expansion during training, transfers spatial representations from established to new stations using attention-based alignment, and incorporates income-stratified regularization. Evaluated on bike-sharing networks in New York City and Seattle, the model achieves state-of-the-art predictive accuracy across various expansion scenarios.

Why it matters — It prevents predictive models from perpetuating historical infrastructure inequalities, allowing planners to identify latent demand in low-income neighborhoods and allocate new stations more equitably without sacrificing overall system efficiency.

arXiv · pdf

preprint

Huy Nguyen, Yue Lin
arXiv cs.AI

census data · Los Angeles · Chicago

Is Your Neighborhood Safe? Place-based Stigma in Large Language Models' Urban Safety Judgments

The study evaluated seven instruct-tuned large language models on their safety ratings of 186 neighborhoods in Los Angeles and Chicago under three conditions: coordinates-only, name-only, and name+coordinates. The ratings were compared against violent crime statistics and American Community Survey demographic data. The models showed flat safety ratings when given coordinates alone, whereas neighborhood names drove most of the variation and lowered safety ratings in proportion to the share of Black residents in Chicago and Hispanic residents in Los Angeles, even when controlling for actual crime and income levels.

Why it matters — It demonstrates that large language models rely on demographic stereotypes embedded in neighborhood names rather than geographic coordinates to assess urban safety, and that this bias actually increases in models with superior geographic knowledge. This reveals a fundamental trade-off for urban decision-support systems, where removing neighborhood names to eliminate demographic bias simultaneously degrades the models' ability to leverage genuine crime data.

Caveat: The findings are based on safety judgments across two major US cities and may reflect specific American patterns of racial segregation and spatial stigma.

arXiv · pdf

preprint

Hilda Adwubi Osei, Catherine Tenewaa Osei, Desdemona Yaa Asobayire
arXiv cs.LG

Municipal Solid Waste Management · Healthcare and Environmental Waste Management · Wastewater Treatment and Reuse · random forest · remote sensing classification

Learning from waste: Machine Learning for health risk prediction and computer vision-based sorting in Ghana

This study developed two machine learning models using 2022 field data from Atonsu in Kumasi, Ghana. First, a Random Forest classifier predicted illness categories from demographic and waste disposal survey data, achieving a macro F1 score of 0.63 on a test set of 69 respondents and identifying disposal method as the top predictor. Second, a MobileNetV2 computer vision model classified waste images for automated sorting, reaching 88.2% accuracy and an F1 score of 0.87 on a test set of 415 images.

Why it matters — The research provides quantitative validation for a previously unproven link between household waste practices and specific illness patterns in Ghana, while demonstrating that a low-cost, single-camera vision system can effectively sort waste in resource-constrained settings without relying on expensive multi-sensor hardware.

Caveat: The health prediction model was trained and evaluated on a very small sample size of 69 respondents who reported illness.

arXiv · pdf

Still cited

David Harvey (2008), The Right to the City 1 of today's items cite it · 16 of 3240 in the archive stand on it

Also published today

These appeared today in journals we track. Their abstracts are not openly available, so we cannot summarise them.