Abstract: It is clear that since the COVID-19 pandemic started in the Philippines, education is one of the most affected areas. After more than a year of struggling with different community lockdowns and the alarming consistency with the increasing number of confirmed cases each day, students and teachers are now left with the choice to voice out their frustrations, activism, opinions, and ideas regarding online classes through different social networking sites, most especially Twitter. With the influx of tweets available in the internet sphere, the authors of this study decided to conduct a sentiment analysis to categorize the overall opinions of Filipino citizens about the current state of education after more than a year of adapting with the distance learning practices that are now considered as the new normal. The authors utilized rtweet, a built-in package available in R programming to perform opinion mining on Twitter data collected through the package related to online class during pandemic. Through sentiment lexicons available in R such as bing and afinn, the results show that most of the tweets about online learning in the Philippines turned out to be neutral. The positive responses are 55.77% while 44.23% of the sentiments collected are negative. To evaluate the accuracy rates of results, the authors used three classification techniques namely Naïve Bayes, logistic regression, and random forest. Naïve bayes and logistic regression both show 69.23% accuracy rate and random forest calculated 71.15% accuracy in identifying whether the given tweet is a positive, negative, or neutral sentiment.