About this project
To predict crime and to stop it before it takes place, is this science fiction or reality? Making it a reality is something both researchers and the police have been working on for the past decades. If you are able to predict when and where a crime will take place, it will give the police valuable information. Information that can lead to the prevention of the crime. However, if you want to be able to prevent a crime, you first need to make a correct assessment, a correct prediction. In the GeoPred-study we aim to predict different types of violent- and property crimes in Malmö with the aid of different geographical prediction models. Then we compare how accurate and reliable the prediction models are over time. We are also investigating if it is possible to use aggregated results from the prediction models, to predict levels of fear of crime in an area. A few of the questions we aim to answer are the following:
- In a comparison what prediction model will render more accurate and reliable predictions of different types of violent and property crimes in Malmö?
- How many crimes can a prediction model predict and does this differ depended on crime type?
- For how long time does the predictions offer value?
- Is it possible to use prediction models developed to predict crimes to predict fear of crime in an area?
- Is it crime history or environmental risk factors, if any, which offer the best predictions of crime and fear of crime?
- How much information do you really need to make good predictions of crime and potentially fear of crime?
- Is the prediction model that potentially renders the best predictions practically feasible to use, on a daily basis for the police?