InclusiViz: Visual Analytics of Human Mobility Data for Understanding and Mitigating Urban Segregation

TL;DR

The paper addresses urban segregation across activity spaces by leveraging human mobility data and introduces InclusiViz, a visual analytics system with a three-level socio-spatial workflow, a Deep Gravity–based mobility model enhanced with SHAP explanations, and a What-If feedback loop for evaluating interventions. It integrates mobility, sociodemographic, and POI data to compute segregation metrics (e.g., SI and BI), detect mobility communities with Leiden, and prioritize CBGs for action using TOPSIS, all presented through four coordinated views including a Dorling map and a Filter Bubble Glyph. The approach demonstrates superior predictive performance (DG+S+V across CPC, JSD, Pearson, RMSE, NRMSE) and enables real-time, interpretable what-if analyses validated by two Houston case studies and expert interviews. The work offers a practical, open-source platform for social scientists and urban planners to design targeted, data-driven interventions that promote inclusive city spaces.

Abstract

Urban segregation refers to the physical and social division of people, often driving inequalities within cities and exacerbating socioeconomic and racial tensions. While most studies focus on residential spaces, they often neglect segregation across "activity spaces" where people work, socialize, and engage in leisure. Human mobility data offers new opportunities to analyze broader segregation patterns, encompassing both residential and activity spaces, but challenges existing methods in capturing the complexity and local nuances of urban segregation. This work introduces InclusiViz, a novel visual analytics system for multi-level analysis of urban segregation, facilitating the development of targeted, data-driven interventions. Specifically, we developed a deep learning model to predict mobility patterns across social groups using environmental features, augmented with explainable AI to reveal how these features influence segregation. The system integrates innovative visualizations that allow users to explore segregation patterns from broad overviews to fine-grained detail and evaluate urban planning interventions with real-time feedback. We conducted a quantitative evaluation to validate the model's accuracy and efficiency. Two case studies and expert interviews with social scientists and urban analysts demonstrated the system's effectiveness, highlighting its potential to guide urban planning toward more inclusive cities.

Figures (8)

• Figure 1: System overview - InclusiViz comprises three modules: Data, Analysis Pipeline, and Visualization, to support interactive segregation analysis (Comm and CBG) and iterative intervention development (POI).
• Figure 2: In the data analysis pipeline, the Segregation Measures (A) computes each CBG's Segregation Index (1) and Bridging Index (2), and then ranks them using Dual-Attribute Ranking (3). The Mobility Data Modeling (B) trains a Deep Learning Model for a social group (1), enhanced with Feature Impact (2) and What-If Analysis (3) modules.
• Figure 3: The interface of InclusiViz guides experts through a segregation analysis and iterative intervention design workflow. In the Community View (A), experts analyze the sociodemographic profiles and interconnections of mobility-based communities, identifying areas with potential segregation. Experts then drill down into the CBG View (B) to select a target CBG for intervention, investigating how its features influence inflow patterns from different social groups. Meanwhile, the Map View (C) offers geographic context for these mobility patterns. Finally, experts use the What-If View (D) to simulate urban planning interventions, with the predicted impact of those changes reflected directly in the CBG View (B) and the Map View (C).
• Figure 4: (A) The income Community Signature design to show the income distributions of CBGs in community B. (B) shows an alternative design: traditional boxplots. (C) The visual encoding of a cell in the Flow Matrix.
• Figure 5: (A) The Filter Bubble Glyph design to demonstrate one CBG's resident sociodemographics and flow patterns to target CBG. (B) and (C) show two alternative glyph designs.