The use of information technology is growing rapidly marked by public
opinion that can be conveyed indefinitely through social media. Microblogging
services such as Twitter allow users to express opinions, feelings, experiences and
other things that concern them. The topics discussed by the public on Twitter also
vary, including the Pertukaran Mahasiswa Merdeka program which is one of the
Merdeka Belajar Kampus Merdeka (MBKM) programs. This program was made
with the aim of one of them being to study across campuses (domestic and
overseas), so that cross-cultural and ethnic brotherhoods are built, with a credit
transfer system of 20 credits or for 1 semester. This program raises sentiments in
the community, the large number of opinions that require classification according
to sentiments owned so that it is easy to get opinion trends towards the Pertukaran
Mahasiswa Merdeka program whether they tend to have positive or negative
opinions. In conducting the analysis, the data was obtained from the scraping
process using the Python programming language. Data from the results of scraping
were obtained as many as 11,319 which were then pre-processed by going through
the stages of cleansing, case folding, tokenizing, normalization, filtering, and
stemming. The data labeling process was carried out using Lexicon Senticnet 7 by
obtaining a total of 7.107 positive sentiment data and 1.850 negative sentiment
data. The classification method used is the Naïve Bayes Classifier Algorithm and
obtains an Accuracy value of 82%. Confusion Matrix results get the Precision score
of 84%, the Recall of 95% and the F1-Score of 89%.