The Covid-19 virus has become a global pandemic, including Indonesia. Various efforts have been made by the government to reduce the negative impact of this pandemic, one of which is through the launch of the Peduli Lindungi application. This research was conducted to classify the public sentiment towards the Peduli Lindungi application based on the review data on the Google Play site. The data collection period starts from July 1 to November 30, 2021. This period was chosen because the government requires the public to participate in preventing the spread of Covid-19 through the Peduli Lindungi application. Initial data labeling uses the Textblob library. The sentiment analysis is classified using the Support Vector Machine algorithm based on two classes, namely positive and negative. The results showed that the classification accuracy performance reaching 81.35%. The Sentiment analysis results are sufficient to display public opinion. The association of words on negative sentiment indicates that users are less satisfied with the performance of the Peduli Lindungi application. Especially in terms of applications, vaccines, certificates, locations, check-ins, registers, and service complaints. Thus, the results of this study are expected to be able to provide information to Peduli Lindungi to focus on improving applications based on complaints written by users and it is hoped that decision makers can improve the performance of Peduli Lindungi applications based on this opinion.The Covid-19 virus has become a global pandemic, including Indonesia. Various efforts have been made by the government to reduce the negative impact of this pandemic, one of which is through the launch of the Peduli Lindungi application. This research was conducted to classify the public sentiment towards the Peduli Lindungi application based on the review data on the Google Play site. The data collection period starts from July 1 to November 30, 2021. This period was chosen because the government requires the public to participate in preventing the spread of Covid-19 through the Peduli Lindungi application. Initial data labeling uses the Textblob library. The sentiment analysis is classified using the Support Vector Machine algorithm based on two classes, namely positive and negative. The results showed that the classification accuracy performance reaching 81.35%. The Sentiment analysis results are sufficient to display public opinion. The association of words on negative sentiment indicates that users are less satisfied with the performance of the Peduli Lindungi application. Especially in terms of applications, vaccines, certificates, locations, check-ins, registers, and service complaints. Thus, the results of this study are expected to be able to provide information to Peduli Lindungi to focus on improving applications based on complaints written by users and it is hoped that decision makers can improve the performance of Peduli Lindungi applications based on this opinion.