Sentiment Analysis of the Issue of Eliminating the Independent Curriculum using the Naïve Bayes Classifier Algorithm

Abstract

Sentiment analysis regarding the issue of eliminating the Independent Curriculum is a crucial tool for understanding public opinion, particularly among students and teachers, on changes in the education system. This study applies the Naïve Bayes Classifier method to classify positive and negative sentiments from data collected through social media platforms such as YouTube. The collected data undergoes text preprocessing techniques, including cleaning, case folding, tokenization, stopword removal, and stemming, to enhance model accuracy. The analysis is conducted using Python 3.12.3 in Google Colab with the Naïve Bayes Classifier algorithm. The results demonstrate strong performance, with a positive precision of 5% and a recall of 68%, using an 80% training and 20% testing data ratio. Findings indicate that the overall sentiment leans more negative than positive, with the majority of respondents supporting the elimination of the Independent Curriculum. This study validates the effectiveness of the Naïve Bayes method in sentiment analysis and highlights the importance of text preprocessing in improving model accuracy. Furthermore, there is potential for exploring other methods, such as word embedding and deep learning, to enhance model performance. The findings of this study can serve as a valuable reference for policymakers in understanding public opinion before making further decisions in the education sector.

Keywords

• independent curriculum, naïve bayes classifier, sentiment analysis