Building Unstructured Crime Data Prediction Model (Practical Approach)
Mona Mowafy, Amira Abdallah Elsayed Rezk, Hazem Mokhtar El-Bakry · INTERNATIONAL JOURNAL OF COMPUTER APPLICATION · 2018
In the most of police stations, it's essential to classify the crimes, according to their types before applying the methods of pattern identification on the crime data.Also the legal organizations aim to divide the crimes into categories for various purposes related to the courts' procedure, such as assigning different types of court to different type of crimes.As a result, the crime type prediction is a decisive part of a crime analysis.There is a little research, in methods and techniques that can predict the crime type from unstructured text.This paper, focuses on how practically building an unstructured crime data prediction model using the scikit-learn Python Toolkit.The main goal of the constructed model is predicting the type of the crime, according to the unstructured data of the crime incident reports for improving the policeman decisions and reduce their investigation efforts..