Klasifikasi Cerita Pendek Berbahasa Bali Berdasarkan Umur Pembaca dengan Metode Naive Bayes

Abstract

Short stories are short stories that tell an event that has happened in a short and clear way. Parents should be able to choose short stories that are suitable for their children because if the stories that parents bring to children are not in accordance with their age, it can affect the development of children. In this study, we will build a system that can classify text. The method used in this research is Nave Bayes with feature selection, namely Genetic Algorithm. This research was conducted to help parents so that their children do not read short stories that are not appropriate for their age so that they do not interfere with their child's development. The data used are children's short stories, youth short stories and adult short stories in Balinese. The best model performance is generated in the training and validation process using new data. The results of testing the Naïve Bayes method without feature selection are 66% accuracy, 66% precision, 67% recall and 66% F1-score. While the Naïve Bayes method uses feature selection, namely 72% accuracy, 72% precision, 78% recall and 73% F1-score.