Analisis Sentimen Pengguna Twiter terhadap Perubahan Kebijakan Skripsi sebagai Syarat Wajib Kelulusan menggunakan Metode Naïve Bayes Classifier

Abstract

The Minister of Education, Culture, Research, and Technology, Nadiem Makarim, has issued a policy to abolish theses, dissertations, or final papers as mandatory graduation requirements for undergraduate and postgraduate students in universities. The requirement to write a thesis is still enforced in most universities in Indonesia to obtain a bachelor's degree. The advancement of information system technology and the ease of accessing social media have caused news to spread rapidly. This policy has sparked pros and cons among the public, including on the social media platform X (formerly Twitter). Some people agree with it, considering that it can reduce the burden on students and increase the relevance of higher education to the needs of the job market. However, others argue that abolishing theses could lower the quality of university graduates and that the replacement could be even more burdensome. The purpose of this research is to understand Twitter users' sentiments towards the policy of abolishing theses as a graduation requirement and to determine the accuracy of the Naïve Bayes Classifier in classifying these sentiments. The data used consists of 656 tweets, which were processed through several stages, including cleaning, case folding, normalization, stopword removal, tokenizing, and stemming. The data was then labeled using a lexicon-based approach, resulting in 353 negative labels and 273 positive labels. The data was subsequently weighted using TF-IDF for the classification process. The dataset was split into training and testing data with a ratio of 90:10. After classification, the study found that the Naïve Bayes Classifier successfully categorized sentiments with an accuracy of 76%.