Studi Komparasi Metode Analisis Sentimen Naïve Bayes, SVM, dan Logistic Regression Pada Piala Dunia 2022

Abstract

The world cup is the most popular sporting event in the world. The 2022 World Cup will be held for the first time in the Middle East, in the country of Qatar to be precise. Its implementation was colored by various controversies ranging from human rights issues, LGBT+ issues, issues of alcoholic beverages, and so on which were so busy in the mainstream media. Various sentiments and opinions have emerged on social media regarding the implementation of the world cup, some have positive opinions and some have negative ones. Sentiment analysis was carried out to find out the main opinions that are developing in society regarding the 2022 world cup, the results can then be used as input and consideration for policy makers. This study uses the snscrape library running on the Python programming language to collect tweets related to the 2022 World Cup on the Twitter social media platform on the first day of the World Cup. The collected data then enters the pre-processing, splitting, TF-IDF stage, before it is ready to be used for modeling. The method used in this research is Bernouli Naïve Bayes, Support Vector Machine, and Logistic Regression. The evaluation results show that the Bernouli Naïve Bayes method produces a precision parameter value of 71%, a recall parameter of 99%, and an accuracy of 76%. While the Support Vector Classifier method produces precision parameter values of 94%, 93% recall parameters, and 92% accuracy. The Logistic Regression method produces a precision parameter value of 93%, a recall parameter of 93%, and an accuracy of 92%.