Font Size: 

The government provides a cash transfer scheme (BLT) to people affected by covid-19. Through social media Twitter, people have given their opinion of positive, neutral, and negative tones regarding the BLT program. The purpose of the study is to classify public opinion on the BLT program written on social media Twitter. The method used is natural language processing text analysis using a machine learning approach with the Support Vector Machine algorithm. Data from Twitter is retrieved using Twint tools. The total data obtained was 2170 tweets with the keywords "bantuan sosial", "blt", and "bantuan corona" during the covid-19 pandemic. Vader tools and Google translate dictionaries are used to reduce human intervention in data labelling. The highest accuracy result of the testing dataset of 82.5% was obtained with the implementation of the SVM parameter: Kernel = RBF, C = 3, gamma = 1 and the value of k = 27. This study contributes to knowledge about the opinions formed in the community for one of the government programs, namely the covid-19 BLT program

Sentiment analysis; Cash transfer scheme; Covid-19; Support Vector Machine; Twitter