Analisis Sentimen Opini Publik Terhadap Pekan Olahraga Nasional Pada Instagram Menggunakan Metode Naïve Bayes Classififer

Abstract

National Sports Week (Pekan Olahraga Nasional) held in October 2021 in Papua has brought many pros and cons from the public. This topic allows public for give criticism, suggestions, and opinions about the National Sport Week 2021. Instagram in one of social media that popular place for deliver public opinion. The process of sentiment analysis can find and solve problems based on public opinion on social media such as Instagram. The classification method used in this research is Naïve Bayes Classifier. The dataset can be obtained from data crawling process using the google chrome extension: IGCommentExport. The data the labelled as positive, neutral, or negative. The labelling process result showed 965 negative data, 256 neutral data, and 770 positive data. Then pre-processing is carried out on the data that has been labeled before, also word weighting process using TF-IDF. After that modelling is carried out using Naïve Bayes Classifier and the last process is evaluation-testing. The high accuracy of the result from fourth experiment which compare 90% data training with 10% data testing produce 75% accuracy. While the result of sentiment test show that negative sentiment more than positive sentiment and neutral sentiment.