Sentiment Analysis on Social Media (Twitter) about Vaccine-19 Using Support Vector Machine Algorithm

Abstract

Currently the world is experiencing a Corona Virus Disease (Covid-19) pandemic which attacks the respiratory tract and spreads very quickly to various countries including Indonesia, so the World Health Organization (WHO) has declared Covid-19 as a pandemic. To overcome this pandemic, experts in the medical field also intervened by making vaccinations to strengthen human immunity against the Covid virus. This sentiment analysis was carried out to see opinions on the object, namely the existence of a Covid-19 vaccine. Data collection by crawling data with the keyword ‘Covid Vaccine’. The method that will be used is the Support Vector Machine (SVM). The analysis was carried out by comparing the classification accuracy values of the two SVM kernel functions, namely linear and Radial Basic Function (RBF). The results of the study obtained positive sentiment of 43.5%, negative of 19.1%, and neutral of 37.4%. Then the evaluation of the system using the confusion matrix obtained an accuracy value for the linear kernel of 79.15%, a precision value of 77.31%, and a recall value of 78.09%. While the RBF kernel has an accuracy of 84.25%, a precision value of 83.67%, and a recall value of 81.99%. While the cross validation obtained the optimum value at $\mathrm{k}=1$ with an accuracy value of 80.18% for the linear kernel and 85.88% for the RBF kernel. So the RBF kernel has a higher accuracy than the linear kernel.