Clustering Crime Theft : A Datamining in Urban Areas
Emilya Ully Artha, Aditya Arif Nugroho, Agus Setiawan, Endah Ratna Arumi, Ardhin Primadewi, Setiya Nugroho, Sunarni

Universitas Muhammadiyah Magelang

Abstract

In the Temanggung District Court, case data handled is only stored without further analysis. In the last three years, the types of cases that are often handled are criminal acts of theft. The case of criminal theft in 2013 is 37 cases, 2014 is 54 cases and 2015 is 83 cases. Based on these data, there is no further analysis on the characteristics of the tendency of theft and prone areas of criminal acts of theft and related matters such as the level of education, occupation and age of the perpetrator and the place of the incident. It affects law enforcement agencies have not been able to determine the strategic steps to reduce the theft rate in the jurisdiction of Temanggung Regency. Proposed a cluster-labeling strategy based on a combination of clustering evaluation techniques. They consider the compactness of the corresponding clusters and the separation between them ant the principal parameters which distinguish between ^normal^ and ^abnormal^ behavior in the analyzed network.
To overcome this problem, K-Means Algorithm clustering method can be applied to analyze data of theft case which handled in Temanggung Regency area in 2013-2015. The use of this technique is expected to provide knowledge previously hidden in the data warehouse so it becomes valuable information. With the method of K-Means clustering algorithm is expected to be known areas that are prone to criminal theft by comparing the level of vulnerability in each region.

Keywords: data mining- k-means- clustering- criminal theft- data

Topic: Other Related Topics