Albanian News Category Predictor System using a Multinomial Naïve Bayes and Logistic Regression Algorithms

In this paper, we proposed a system for the classification of different news in the Albanian language using the Multinomial Naïve Bayes and Logistic Regression algorithms. The system is web-based and developed with Python programing language and Flask web framework. For the classification of Albanian language news, we have used 70% of the dataset for training and 30% for tests. The result shows that our system can use for the classification of Albanian language news. The proposed Multinomial Naïve Bayes and Logistic Regression classifiers were analyzed and compared for the following evaluation parameters: Accuracy, Precision, Recall, F1-score, and Confusion Matrix. The obtained results showed that the Logistic Regression algorithm has higher accuracy compared to the Naïve Bayes Multinomial algorithm in the classification of news in the Albanian language.