Laura A. Bautista-Montejano

Laura Bautista-Montejano is a Ph.D. candidate in the Urbanism Graduate Program at the Universidad Nacional Autónoma de México (UNAM). She holds a Master’s degree in Spatial Planning from the Center for Research in Geospatial Information Sciences (CentroGeo).
Supervised by Dr. Lucía Carmina Jasso López at UNAM's Institute for Social Research, her doctoral thesis investigates the spatial concentration of crime at micro-urban scales, with a specific focus on street segments. Her broader research sits at the intersection of evidence-based urban planning, multidisciplinary spatial analysis, and walkability research. Through her work, she aims to leverage spatial data to better understand urban safety, inform public policy, and support the development of safer, more accessible, and walkable environment systems in Mexican cities.


Sesión

12/11
12:20
15minutos
Visualizing Street Robbery of Pedestrians in Street Networks in Iztapalapa Borough: An Experience Using Network Kernel Density Estimation in Python
Laura A. Bautista-Montejano

"That crime is spatially concentrated now seems incontestable" (Frith et al., 2017, p. 344). This statement powerfully summarizes why the use of maps has become such a common resource for identifying crime concentrations. This research draws on that premise to analyze, specifically, the spatial concentration of street robbery of pedestrians.
One of the most common methods for representing hotspots across many disciplines is the use of heat maps (Chainey et al., 2008; Pokojski et al., 2021; Ratcliffe, 2010). The most frequent approach (when data is available) is using the Kernel Density Estimation (KDE) technique. This technique estimates gradients from lower to higher concentration based on point geometry patterns. The final representation achieves a smoothed visual effect, which can be highly appealing and makes it possible to quickly locate hotspots (Słomska-Przech et al., 2021).
However, using this technique to represent crime in urban environments can be misleading, mainly because the urban fabric does not necessarily operate according to the properties of continuous space. The KDE technique belongs to planar spatial analysis and follows a geometric orientation. By contrast, topological analysis adds streets represented as nodes and links, so that unlike the geometric orientation (where points are adjacent), in network space they may not be (Okabe & Sugihara, 2012).
The topological analysis of the road network exhibits a scaling pattern characterized by nonlinear relationships (Jiang & Okabe, 2014), where cause and effect do not maintain a direct proportion. In the geographic domain, this "nonlinearity" reflects the fractal nature of urban forms (complex, far from equilibrium systems), where the number of small elements far exceeds that of large ones (Jiang, 2015).
The technique explored in this research to visualize point density over a street network is known as "Network Kernel Density Estimation," or NKDE. Unlike KDE, NKDE (being one-dimensional) does not tend to over-detect point patterns (Xie & Yan, 2008). Taking this into account, for the exploratory spatial analysis of crime, using the NKDE technique is preferable to other hotspot detection methods. However, technical and technological challenges remain regarding the widespread use of street-network visualizations.
The objective of this research is to present a methodological strategy for visualizing the concentration of street robbery of pedestrians using the NKDE technique. The challenges and opportunities associated with the options available through open source software will be discussed. This discussion is illustrated through an applied case: street robbery of pedestrians reported in Iztapalapa borough between 2021 and 2025, using data from the Mexico City Attorney General's Office (Fiscalía General de Justicia de la Ciudad de México), processed using the open-source library PyNKDV (Py Network Kernel Density Visualization; Chan et al., 2023), which implements the ADA/HA algorithm (Aggregate Distance Augmentation/Hybrid Augmentation), the most efficient in terms of computation time. However, users must take certain technical and methodological precautions when applying it. These considerations, among other specific technical issues (such as how to obtain spatial data), are expected to be discussed during the presentation.

References
Chainey, S., Tompson, L., & Uhlig, S. (2008). The Utility of Hotspot Mapping for Predicting Spatial Patterns of Crime. Security Journal, 21(1–2). https://doi.org/10.1057/palgrave.sj.8350066

Chan, T. N., Zang, R., Ip, P. L., U, L. H., & Xu, J. (2023). PyNKDV: An Efficient Network Kernel Density Visualization Library for Geospatial Analytic Systems. Proceedings of ACM Conference on Management of Data (SIGMOD), 23. https://doi.org/10.1145/3555041.3589711

Frith, M. J., Johnson, S. D., & Fry, H. M. (2017). ROLE OF THE STREET NETWORK IN BURGLARS’ SPATIAL DECISION-MAKING*. Criminology, 55(2). https://doi.org/10.1111/1745-9125.12133

Jiang, B. (2015). Geospatial analysis requires a different way of thinking: the problem of spatial heterogeneity. GeoJournal, 80(1), 1–13. https://doi.org/10.1007/s10708-014-9537-y

Jiang, B., & Okabe, A. (2014). Different ways of thinking about street networks and spatial analysis. Geographical Analysis, 46(4), 341–345. https://doi.org/10.1111/gean.12060

Okabe, A., & Sugihara, K. (2012). Spatial Analysis along Networks: Statistical and Computational Methods. In Spatial Analysis along Networks: Statistical and Computational Methods. https://doi.org/10.1002/9781119967101

Pokojski, W., Panecki, T., & Słomska-Przech, K. (2021). Cartographic visualization of density: exploring the opportunities and constraints of Heat Maps. Polish Cartographical Review, 53, 21–36.

Ratcliffe, J. (2010). Crime Mapping: Spatial and Temporal Challenges. In Handbook of Quantitative Criminology. https://doi.org/10.1007/978-0-387-77650-7_2

Słomska-Przech, K., Panecki, T., & Pokojski, W. (2021). Heat maps: Perfect maps for quick reading? Comparing usability of heat maps with different levels of generalization. ISPRS International Journal of Geo-Information, 10(8), 562. https://doi.org/10.3390/IJGI10080562/S1

Xie, Z., & Yan, J. (2008). Kernel Density Estimation of traffic accidents in a network space. Computers, Environment and Urban Systems, 32(5), 396–406. https://doi.org/10.1016/j.compenvurbsys.2008.05.001

Sistemas de Información Geográfica, Cartografía y Geoinformática
Multifuncional (Remoto)