Investigating the Effectiveness of Deep Learning and Machine Learning for Bangla Poems Genre Classification
Rayhanuzzaman Rayhanuzzaman, Tanjim Mahmud, Utpol Kanti Das, Sultana Rokeya Naher, Mohammad Shahadat Hossain, Karl Andersson
Abstract
Bangla Poems Genre Classification is crucial for understanding and preserving cultural heritage. In this paper, we employ both machine learning and deep learning approaches to classify Bangla poetry genres. Our study encompasses a range of models, including Multinomial Naive Bayes, Support Vector Machine, Random Forest, Logistic Regression, Decision Tree, Extreme Gradient Boosting, CNN, GRU, and BERT. Experimental results reveal that the Naive Bayes model achieves the highest accuracy of 77%, outperforming other models. Additionally, we present performance metrics for each model, providing a comprehensive evaluation. This research contributes to the advancement of regional language natural language processing and fosters a deeper understanding of Bangla poetry through computational analysis.