Data Analysis Analisis Sentimen Publik Terhadap Sekolah Tatap Muka Saat Covid-19 Pada Twitter Menggunakan Metode Lexicon Based

Abstract

Lately the whole world is rampant hit by the Corona Virus outbreak. Indonesia is not spared from the spread of this virus, with the Covid-19 in Indonesia making many adverse impacts that arise such as in the social, tourism, economic and education fields. The Minister of Education and Culture issued a circular on March 24, 2020 which contains about the learning/teaching process will be done online or online to reduce the number of virus spread in schools, after online school trials there are still many shortcomings, for example inadequate internet access. Because it was felt that online learning was less effective, of course this policy invited a lot of public comments, especially on social media twitter. This study aims to find out the comments whether it falls into the classification of sentiments that have been dividen into 5 classes, namely very positive sentiment, positive sentiment, negative sentiment, very negative sentiment, and neutral as well as to know the percentage results of each class. Lexicon Based's research method uses vader sentiment. The percentage accuracy results of 3000 tweet data were 1.3% very positive, 6.04% positive, 3.9% negative, 0.54% very negative, and 88.23% neutral. Keywords: sentiment analysis, lexicon based, covid-19, pandemic, twitter