Analisis Sentimen Pembelajaran Daring Pada Masa Pandemi Covid-19 di Indonesia

Abstract

The spread of COVID-19 is designated as a pandemic that has spread to the entire population in the world and one of the affected people is Indonesia. In reducing the spread of COVID-19, the government issued rules to minimize positive numbers from the pandemic. The rules set include large-scale social restrictions which in the world of education also apply these rules. In the education system, it is made into an online system which was previously a face-to-face system. With the issuance of the rules for this online system, there are pros and cons in responding to it. There are people who object to requiring an internet connection to go through the learning process which will cost more for this. Regarding these problems, news about the application of online learning systems is crowded on social media, one of which is YouTube. The discussion about this online system makes the public have an opinion in the comment column provided by the social media. Various public opinions have emerged, so this research can assess a sentiment contained in social media regarding this online learning system. The research tries to use the nave Bayes method and the LDA sequential deep learning algorithm which is still in the experimental stage by setting the results manually but according to the procedures of each method. The nave Bayes method found negative comments of 65.33% and positive values ​​of 23.33% with an accuracy value of 56.45%. Meanwhile, using the LDA sequential deep learning algorithm method, 35 positive comments, 95 negative comments, and 17 neutral comments were obtained from 150 data taken from 5 videos on YouTube. Judging from the results with the method that has been applied, what is of great value is the negative value of the discussion.