A Comparative Analysis of Bangla Crime News Categorization Using Most Prominent Machine Learning Algorithms

Abstract

This work is dedicated to Bangla Crime Type Classification. Despite several comprehensive crime textual datasets are available on different languages but there is no available dataset on Bengali language. In this work, we created a dataset of Bangla crime articles from different news portals like (Prothom Alo, Jugantor, Noya Digonto), which contains around 3,500 articles. Then we have built our crime classifier model and trained the classifier with our own dataset. This paper explores the use of machine learning approaches, or more specifically, four supervised learning Methods, namely Logistic Regression, Support Vector Machine (SVM), K-Nearest Neighbour (KNN) and Naive Bayes (NB) for categorization of Bangla Crime News Articles. Finally we have summarized the experimental result in tabular form. We can see that significant improved accuracy can be achieved using machine learning algorithms in classifying Bangla Crime data. The final experimental result shows that proposed model is able to achieve around 87% accuracy.