Derek Vildosola

Bio: Derek Vildosola is an academic researcher from University of Miami. The author has contributed to research in topics: Routine activity theory & Crime prevention. The author has an hindex of 2, co-authored 2 publications receiving 10 citations.

Crime in an Affluent City: Spatial Patterns of Property Crime in Coral Gables, Florida

Abstract: This study investigates the spatial and temporal patterns of property crime at the census block level in the mid-sized, medium- to high-income city of Coral Gables, Florida, USA, between 2004 and 2016. Specifically, we analyzed residential and vehicular burglary. We used Emerging Hot Spot Analysis (EHSA) to locate and identify crime hot and cold spots over time. In order to understand the role that various sociodemographic variables play in predicting crime patterns, geographically weighted regression (GWR) was used to analyze the spatial clustering of crime, and commercial areas, renter percentage, median household income, and multifamily households. Our results revealed consistent hotspots for residential and vehicle burglary within the northeast area of the city, while vehicle burglary had hotspots along U.S. Route 1 (US-1)—a main road in Coral Gables—and around the University of Miami, with emerging hotspots within the northwest part of the city bordering lower-income areas. Hotspots were associated with structural factors within and around the city including more multifamily homes, higher poverty rates, more renters, and greater economic disadvantage in surrounding municipalities. Social disorganization and routine activity perspectives are supported as frameworks to understand crime patterns in this context. The findings suggest that policymakers should target specific locations using geospatial analyses to better address property crime.

Crime in an Affluent City: Applications of Risk Terrain Modeling for Residential and Vehicle Burglary in Coral Gables, Florida, 2004–2016

Abstract: Risk Terrain Modeling (RTM) — an innovative geospatial approach to analyze the locations of high crime areas within cities— was used to analyze criminogenic spaces and identify the riskiest places contributing to vehicle and residential burglary in the city of Coral Gables, Florida from 2004 to 2016. Official crime incident data on residential and vehicle burglary were provided by the Coral Gables Police Department. We investigated the role of environmental predictors of crime by analyzing the effects of the designated riskiest places including alcohol vendors, car dealers, gas stations, bars, secondary/post-secondary schools, grocery stores, and restaurants. We identified risky places and their proximity to the occurrence of residential and vehicle burglary with a regression process using the RTM technique to determine the Relative Risk Values (i.e., weighted risk) that each risk factor had. We examined different temporal designations, including day and night, day of the week, and monthly intervals. Our results indicated that restaurants and grocery stores located in the downtown area and along the US 1 highway were higher risk locations for the occurrence of both types of crime throughout the city. The locations of risky places in the study area were spatially consistent with high crime areas, supporting our main hypothesis. The use of RTM with targeted community policing strategies could move policy from a reactionary approach to more proactive solutions to crime prevention. Concepts drawn from routine activity theory, as well as crime and place perspectives, both receive moderate empirical support for vehicle and residential burglary outcomes.

Crime in an Affluent City: Applications of Risk Terrain Modeling for Residential and Vehicle Burglary in Coral Gables, Florida, 2004–2016

Abstract: Risk Terrain Modeling (RTM) — an innovative geospatial approach to analyze the locations of high crime areas within cities— was used to analyze criminogenic spaces and identify the riskiest places contributing to vehicle and residential burglary in the city of Coral Gables, Florida from 2004 to 2016. Official crime incident data on residential and vehicle burglary were provided by the Coral Gables Police Department. We investigated the role of environmental predictors of crime by analyzing the effects of the designated riskiest places including alcohol vendors, car dealers, gas stations, bars, secondary/post-secondary schools, grocery stores, and restaurants. We identified risky places and their proximity to the occurrence of residential and vehicle burglary with a regression process using the RTM technique to determine the Relative Risk Values (i.e., weighted risk) that each risk factor had. We examined different temporal designations, including day and night, day of the week, and monthly intervals. Our results indicated that restaurants and grocery stores located in the downtown area and along the US 1 highway were higher risk locations for the occurrence of both types of crime throughout the city. The locations of risky places in the study area were spatially consistent with high crime areas, supporting our main hypothesis. The use of RTM with targeted community policing strategies could move policy from a reactionary approach to more proactive solutions to crime prevention. Concepts drawn from routine activity theory, as well as crime and place perspectives, both receive moderate empirical support for vehicle and residential burglary outcomes.

A multi‐dimensional crime spatial pattern analysis and prediction model based on classification

Abstract: Decades of study have firmly established that crime shows geographical (ie, spatial) patterns [1]. Analysis of spatial patterns is a standard research approach in criminology, just as it is in ecology, epidemiology, and other fields. Spatial patterns may have different dimensionalities, as they can involve points, lines, or areas; they may also vary with resolution. Crime-pattern analysis may be conducted at the level of census tracts, zip-code units, street segments, counties, states, or countries. In this work, after considering a number of possible resolutions, we find and utilize one that seems optimal for crime prediction. Spatial pattern analysis can be density-based (area-based) or distance-based. However, Euclidean distance is not always useful in identifying urban crime patterns: Places that are close together on a map (in terms of Euclidean distance) may in fact be very isolated from each other if they are not joined by streets, are on opposite sides of a river with few bridges, or are in neighborhoods separated by some invisible economic or social barrier that keeps residents apart. On the other hand, densityor area-based spatial pattern analysis seems to fit naturally with the intuitive concept that cities are built up of neighborhoods. Density-based analysis can be further categorized as global or local. The first considers the ratio of observed crime events to the area of the region under study; the latter measures crime incidence for different units within that region. The spatial pattern is only one aspect of the distribution of crime; there are also temporal patterns. Many researchers have studied variation in crime rates between day and night, weekday and weekend, or among different seasons of the Received: 5 December 2019 | Revised: 21 April 2020 | Accepted: 6 May 2020 DOI: 10.4218/etrij.2019-0306