https://padlet.com/danielapriori/hsuebyzd9histpob Homeworks, tasks and mini-tasks en-us 2021-04-21 23:57:00 UTC 2024-07-17 01:48:50 UTC hello@padlet.com https://padlet.net/icons/png/2622.png danielapriori https://padlet.com/danielapriori/hsuebyzd9histpob/wish/1546179827 Risk Terrain Modeling (RTM) is an approach to risk assessment and a novel technique to forecast crime events in relation to their geographical and social context in contrast with classical spatial analysis and statistical techniques based on hotspot mapping that only takes into account how to identify and to tag areas where crime is concentrated. On the other side, this approximation is not only useful in risk assessment about crime but other types of social or public health problems.

Hotspot mapping, the most popular spatial analysis technique to analyze crime events can’t forecast the dynamical risk related to a particular type of crime. Hotspot mapping is based on identifying places with high spatial crime density to direct police action to focus locations in order to suppress crime and deter offenders. The relevant factors in this approximation are onset, recurrence, frequency, desistence, and intermittency, in the context of how these processes influence the concentration of crime. In summary, while hotspot mapping has allowed police to address the concentration of crime, it has generally turned attention away from the social and structural contexts in which crime occurs.

Risk Terrain Modeling (RTM) is an alternative technique mainly directed to face a given risk factor by understanding it as a dynamical function of interaction social, physical, and behavioral factors within a spatial context. As a way of example, in pioneer research made in Newark NJ, social scientists identified environmental factors related to violent crimes by shootings with o without murders. Some factors were locations of drug arrests, locations of gang activity, known home addresses of parolees previously incarcerated for violent crimes and/or violations of drug distribution laws, location of past shooting incidents, and locations of past gun robberies. After this technique was applied to Newark, the town experienced a significant reduction in violent crime. According to department police statistics, overall crime decreased 19% from 2006 through 2009, with murders and shootings decreasing 28% and 40% respectively. Newark the biggest city in the State of New Jersey had an estimated population of over 280000 people in 2009 and the largest municipal police force with 1300 officers. Nevertheless, murders and non-fatal shootings have been increasing from 2000 to 2006. For this reason, the RTM’s success was significant although not enough since in 2008 Newark’s murder rate 23.9 (per 100000 residents) doubled the national average (11.5) for cities with populations greater than 250000 residents.

Risk Terrain Modeling is an effective method that can be developed and extended not only involving machine learning algorithms as predictors between explicative variables and the explained variable but extending its reach to other social problems of interest different from crime risk problems. The algorithm underlining RTM is logistic regression. It consists of assessing the influence of explicative variables in earlier time periods on the occurrence of the explained variable in a subsequent time period. Nevertheless, logistic regression is only one of a lot of possibilities in the world of predictors algorithms. Other alternatives as naive-bayes, support vector machines, k-nearest neighbors, or decision trees can be used in order to assess the occurrence of the explained variable. On the other side, RTM has been extended to social or public health problems. In these cases, the main difficulty is based on the correct choice of risk factors associated with explicative variables. This item is directly related to available databases, information quality, and the layers that can be targeted in a geographic information system.

On summary, Risk Terrain Modeling is a better alternative than other risk assessment methods not only when researchers want to describe one phenomenon in a geographical context, but when they want to relate that phenomenon to social or cultural risk factors. This technique has been proved and has enough empirical support. On the other side is possible to apply it to different areas not necessarily restricted to security. There are very suggestive possibilities to expand the power or reach of the technique using machine learning algorithms in other to forecast risk. 

]]> 2021-05-21 01:24:41 UTC https://padlet.com/danielapriori/hsuebyzd9histpob/wish/1546179827 danielapriori https://padlet.com/danielapriori/hsuebyzd9histpob/wish/1646494429 2021-07-13 00:15:43 UTC https://padlet.com/danielapriori/hsuebyzd9histpob/wish/1646494429