Analisis Sentimen dan Klasifikasi Tweet Terkait Mutasi COVID-19 Menggunakan Metode Naïve Bayes Classifier

Abstract

Towards the end of 2019 in Wuhan City, China, a new type of Corona Virus was discovered which has the scientific name COVID-19 and is a type of virus that causes acute disorders in the human respiratory system. The spread of this virus is very fast and causes mutations of this virus to a more lethal stage than before. Thus, sentiment analysis is expected to be able to determine the trend of public assessment of the COVID-19 mutation. Naïve Bayes Classifier is a method used in research. This method can classify data or opinions into two sentiments, namely positive and negative. The research data comes from Twitter which is taken using the Twitter API with the keyword "covid mutation", for data processing several processes are carried out, namely sentiment classification, data cleaning, and preprocessing so that the final result is obtained. The test results from this study show that the Naïve Bayes Classifier method has an accuracy of 86.67% with an f1-score of 82.00% on positive sentiment and 89.00% on negative sentiment. Based on the results of the study, it can be concluded that the Naïve Bayes Classifier method can be used to analyze sentiment data from tweets about the COVID-19 mutation with an accuracy of 86.67%.

Keywords

• covid-19
• sentiment analysis
• nave bayes classifier
• text mining.