CLASSIFICATION OF GENERAL CRIMINAL CASE TYPES AT THE BENGKULU HIGH PROSECUTOR’S OFFICE USING MACHINE LEARNING ALGORITHMS

Abstract

The handling of general criminal cases at the Bengkulu High Prosecutor’s Office requires an efficient system to classify case types in order to enhance the legal process's effectiveness. One promising approach to optimize this classification process is the application of machine learning algorithms. This study aims to develop a classification model capable of identifying the types of general criminal cases based on the available case data. The algorithms employed in this research include LightGBM and Random Forest, which were evaluated to determine the highest classification accuracy. The dataset consists of case descriptions received by the Bengkulu High Prosecutor’s Office during a specific period, which underwent preprocessing steps such as text cleaning, tokenization, and feature extraction to generate relevant features. The findings reveal that LightGBM outperforms Random Forest in classification accuracy. The resulting model can reliably predict the category of criminal cases, serving as an effective tool to assist the Prosecutor’s Office in automating case grouping. The implementation of this model is expected to improve operational efficiency and support a more transparent and structured case management system.