This Review is on "Irandegani, Z., Mohammadi, R., & Taleai, M. (2019). Investigating Temporal And Spatial
Effects Of Urban Planning Variables On Crime Rate: A Gwr And Ols Based Approach."
In the research, “Investigating temporal and spatial effects of urban planning variables on crime rate,” they looked at how urban planning is a decisive factor when it comes to crime activities. For instance, they investigated the spatial pattern of urban planning in Boston city of Massachusetts which include parameters of land use diversity and number of police station will affect crime activities. The other parameter was the temporal effects of urban planning when considered of their daily and weekly time windows for crime incidents. The base of crime analysis is that crimes do not happen randomly, and which is why we need to perform spatial analysis to protect the order of our society from potential harm by looking into crime patterns. They verified that their crime data has high spatial correlation by looking at the Moran’s I value which is often used in Crime analysis and proves that crime activities happen based on a pattern. Furthermore, crimes are not distributed evenly across space, and time. So, the base of their study was focused on how urban planning variable influence crime activities, in a sense could it be improved in areas where there is high risk? Geographic Weighted Regression Model (GWR) was perfect to highlight the varying effects of urban planning variable across space because space-time behaviour of crime pattern is more apparent on a local level in Urban cities and sometimes reveal a more accurate predication for certain crimes categories (Stein et al, 2019). Urban planning involves land diversity, which is considered different in everywhere around the city, and GWR could captures those relationship that only apply to certain locations.
METHOD
-The author began their research to find clusters of crime activities, but first they had to pre-process the data to eliminate outliers. Afterward, those data will be processed with an algorithm called DBSCAN to determine crime spatial cluster, and then OLS/GWR will be applied to investigate the relationship between extracted crime clusters and independent variables (land use diversity, police station coverage and daily/weekly temporal scales).
- In the second step, when they used DBSCAN, they mentioned that they had to divide the data into n dimension and formed a shape; then counted how many data falls in those area as well within those fixed distance. This approach seems to be similar to nearest neighbour hierarchal spatial clustering that identifies group of incidents that are spatially close within a distance. Lastly, they applied OLS to find the global relationship, and GWR that can change their coefficient accordingly. The independent variable is spatial urban planning parameter that includes number of police station, and land use diversity with a range from 0-1.
-These results suggest that more land use diversity result in an increase in crime activities. On the other hand, more police station results in a decrease in crime activities.
DISCUSSION
-Highest rate of crime occur in areas with low police station. As well, there is positive relationship for land use diversity and crime occurrences. For example, the position of police stations has a significant impact on crime reduction. Generally, the model had helped verified for suitable decisions toward reducing crimes by the positioning of new police stations, increasing the number of police in times with overpopulation, or policymaking based on the land-use diversity in areas with highest land use diversity. But there are also rooms for additional variables such as demographics (i.e., age, income, and gender), which are relevant to the crime rate and not included in this study should be included in future studies. The maps could also be produced in a better quality when they are published. The results not only told us about relationship but those relationship was verified using both type of spatial statistics method. Overall, the discussion told us a lot about how to better divert our resources in areas that needs to be focused on the most.