Automatic Bangla article content categorization using a hybrid deep learning model

Abstract

The process of categorizing or marking documents based on their content is known as document classification or categorization. Bangla is the world's fifth most widely spoken native language and the sixth most widely spoken language overall. Proper methods for automatic document classification are very limited. In this paper, we have introduced a hybrid deep learning model which is the combination of CNN-BiLSTM for automatic article categorization for Bangla language. We have also compared our model with different classifiers to measure its efficiency. Our proposed model acheived 83.48% accuracy.