Renovation of traditional commercial streets needs quantitative prediction of how pedestrian flow distributes across the street system. A 2017 study of one Guangzhou street (Hou & Zheng) showed that combining a space-syntax configurational measure (integration) with a business-type attraction term predicts log pedestrian counts far better than configuration alone, but its regression, its subjective business-attraction weights, and its coefficient ranking (business attraction over spatial integration) were fitted to a single street, and the study found that high intelligibility does not guarantee good prediction because central attractions interfere with movement. The open problem is to turn this single-case result into field knowledge: a transferable, out-of-sample-validated model of pedestrian-flow distribution in traditional commercial streets that combines spatial configuration and business composition, together with a quantified rule for when configuration dominates pedestrian distribution and when commercial attraction dominates. Later work through 2026 has replicated the single-case approach on new streets, pooled centrality–movement associations in a meta-analysis, and documented context-dependence in both directions, but has not established either a validated transferable model or a general dominance rule.

Background

Space syntax is a research program that represents a street or building layout as a graph of the fewest and longest straight lines covering all public space (the axial map) and relates social activity to the graph's configuration. Two standard measures are integration (RnRn or InIn), a depth-based score of how shallow or accessible a line is relative to all others in the system, and connectivity (CnCn), the number of lines a given line directly intersects. Intelligibility is the squared correlation between connectivity and integration across the lines of a system; it expresses how well what a pedestrian sees locally (connections) predicts the line's global importance, and hence how predictable spatial use should be. These measures are computed with the Depthmap software, and pedestrian flow is typically regressed on them at a set of observation spots. The foundational theory of natural movement (Hillier, Penn, Hanson, Grajewski and Xu, 1993) holds that street configuration is the primary generator of pedestrian movement, while attractors such as shops act as multipliers that themselves tend to locate on well-integrated streets — a two-way relationship later formalized as the movement-economy framework and extended by Lerman, Rofè and Omer (2014) with physical-structure factors and the observation that the relative roles of the factors depend on area type. Within this tradition, Hou and Zheng (2017) studied one traditional commercial street in Guangzhou, China, with 26 observation spots counted during three weekend hours. They graded business types at retail frontages into attraction scores JJ of 2, 1, or 0, defined a business-type integration regressor Y=∑iJi/H‾iY=\sum_{i} J_{i}/\overline{H}_{i} (attraction scores weighted by inverse average depth H‾i\overline{H}_{i} of the observation spots), and fitted the regression M=3.111−0.118 In+0.476 YM=3.111-0.118\,In+0.476\,Y on the logarithm MM of pedestrian counts, with reported space-specific coefficients of determination between 0.72 and 0.94. The standardized coefficient of YY (1.407) exceeded that of InIn (−0.527), so business attraction outweighed configuration in this sample, and the authors reported that high intelligibility did not always produce good prediction because central attractive factors interfere with movement. The study's own unresolved issues, which this problem takes up, are: the relation and coefficient ranking rest on one street and few observation spots, so their robustness across traditional commercial streets with different historical forms, tenant mixtures, and urban contexts is unknown; there is no general rule for when spatial configuration dominates pedestrian distribution and when commercial attraction dominates; and the empirical business-type weighting and regression formulation have not been generalized into a transferable, validated model. These are recognized field-level gaps, not only defects of one paper: a 2018 meta-analysis of 14 cross-sectional studies (Sharmin and Kamruzzaman) found centrality measures significantly associated with pedestrian movement on average, but documented associations are not street-level predictive models; a structural-equation study (a path-modeling statistical method that separates direct effects from mediated ones) of Wuhan housing estates (Yang and Vaughan, 2022) found configuration the strongest predictor with significant differences between gated and non-gated area types and called for validation in other contexts; and a study of the historic town of Mymensingh (Aziz, 2018) found the configuration–volume correlation poor because commercial land use dominates there. Recent work on the mutual reinforcement of street networks and functional attractors (Yang and Qian, 2023, on Xi'an) further stresses the circular causality that any dominance rule must confront.

Problem Statement

Establish a validated, transferable quantitative basis for predicting the distribution of pedestrian flow in traditional commercial streets that combines spatial-configuration measures with business-type attraction, and characterize when each factor dominates. Concretely, on the evidence gap left by Hou and Zheng (2017) and left unfilled by subsequent work: (1) determine whether a combined configuration-plus-business-attraction regression of pedestrian counts — such as the integration and business-type integration formulation fitted on one Guangzhou street, or an improved specification of the same two ingredient families — generalizes across traditional commercial streets that differ in historical form, tenant mixture, and urban context, demonstrated by out-of-sample validation at the level of whole streets rather than in-sample fit on the fitting street; (2) derive and test a quantified rule, stated over measurable street-system properties (for example configurational measures, intelligibility, and the spatial arrangement and composition of business types), for the conditions under which spatial configuration dominates pedestrian-flow distribution and the conditions under which commercial attraction dominates — including the reported phenomenon that high intelligibility fails to yield good prediction when central attractions interfere with movement, with the rule calibrated on the fitting streets and tested out-of-sample on held-out streets; and (3) either validate or replace the empirical business-type attraction weighting (the graded attraction scores and depth-weighted summation) with an attraction measure whose mapping is justified against independent evidence of drawing power. A positive resolution requires all three components jointly, and no single component alone resolves the task; their common objective is a predictive model, with quantified accuracy and applicability conditions, that a renovation designer of a traditional commercial street can use before intervention to decide whether to restructure the spatial organization or reorganize the commercial tenant mix. An admissible answer may alternatively show, with a comparably rigorous multi-street design, that no stable transferable combined model or dominance rule exists — that is, quantified heterogeneity across street systems that no measurable street-level covariate explains — which would settle the question negatively for this class of streets; a bare failure of a single fitted model to transfer, where measurable street-level covariates account for the heterogeneity, does not by itself resolve the question negatively.

References

  1. Hou, Xia-na, and Zheng, Li-peng (2017), 'Simulation of Pedestrian Flow in Traditional Commercial Streets Based on Space Syntax', Procedia Engineering 205: 1344-1349, DOI 10.1016/j.proeng.2017.10.117, https://doi.org/10.1016/j.proeng.2017.10.117
  2. Hillier, B., Penn, A., Hanson, J., Grajewski, T., and Xu, J. (1993), 'Natural movement: or, configuration and attraction in urban pedestrian movement', Environment and Planning B: Planning and Design 20(1): 29-66, DOI 10.1068/b200029, https://doi.org/10.1068/b200029
  3. Lerman, Y., Rofè, Y., and Omer, I. (2014), 'Using Space Syntax to Model Pedestrian Movement in Urban Transportation Planning', Geographical Analysis 46(4): 392-410, DOI 10.1111/gean.12063, https://doi.org/10.1111/gean.12063
  4. Sharmin, S., and Kamruzzaman, M. (2018), 'Meta-analysis of the relationships between space syntax measures and pedestrian movement', Transport Reviews 38(4): 524-550, DOI 10.1080/01441647.2017.1365101, https://doi.org/10.1080/01441647.2017.1365101
  5. Aziz, N. (2018), 'Land-Use, Street Configuration and Pedestrian Volume: The Case of a Historic Town, Mymensingh, Bangladesh', Nakhara: Journal of Environmental Design and Planning 15: 137-154, DOI 10.54028/nj201815137154, https://doi.org/10.54028/nj201815137154
  6. Yang, Y., and Vaughan, L. (2022), 'Does area type matter for pedestrian distribution? Testing movement economy theory on gated and non-gated housing estates in Wuhan, China', Computers, Environment and Urban Systems 97: 101868, DOI 10.1016/j.compenvurbsys.2022.101868, https://doi.org/10.1016/j.compenvurbsys.2022.101868
  7. Yang, C., and Qian, Z. (2023), 'Street network or functional attractors? Capturing pedestrian movement patterns and urban form with the integration of space syntax and MCDA', Urban Design International 28(1): 3-18, DOI 10.1057/s41289-022-00178-w, https://doi.org/10.1057/s41289-022-00178-w
  8. Lian, H., and Li, G. (2023), 'Correlation Analysis of Retail Space and Shopping Behavior in a Commercial Street Based on Space Syntax: A Case of Shijiazhuang, China', Buildings 13(11): 2674, DOI 10.3390/buildings13112674, https://doi.org/10.3390/buildings13112674
  9. Zhou, W., Guo, H., Hou, X., Lai, W., and Yao, L. (2024), 'A Statistical Study of the Pedestrian Distribution in a Commercial Wholesale Centre Based on the Traffic Spatial Structure', Buildings 14(6): 1782, DOI 10.3390/buildings14061782, https://doi.org/10.3390/buildings14061782
  10. Wolpert, L., and Omer, I. (2025), 'Comparative analysis of pedestrian volume models: Agent-based models, machine learning methods and multiple regression analysis', Computers, Environment and Urban Systems 117: 102238, DOI 10.1016/j.compenvurbsys.2024.102238, https://doi.org/10.1016/j.compenvurbsys.2024.102238
  11. Al-Naama, A. M., and Abu-Rayash, A. (2026), 'A Comprehensive Review of Space Syntax Applications for Sustainable Urban Development in Commercial Areas', Sustainability 18(10): 5145, DOI 10.3390/su18105145, https://doi.org/10.3390/su18105145
  12. Wang, C., Wang, H., Yao, J., and Chen, Y. (2026), 'Typical Routes and Sequential Movement Patterns in Commercial Spaces', Buildings 16(14): 2777, DOI 10.3390/buildings16142777, https://doi.org/10.3390/buildings16142777