%0 Journal Article
%@ 2686-228X
%A Hasan, Firman Noor
%D 2023
%F repository:28232
%I Forum Kerjasama Pendidikan Tinggi
%J JOSH: Journal of Information System Research
%N 4
%P 1136-1144
%R https://doi.org/10.47065/josh.v4i4.3803
%T (JOSH - S4) - Rizkina, Hasan [2023-07-26]
%U http://repository.uhamka.ac.id/id/eprint/28232/
%V 4
%X The present rate of technological advancement has resulted in the rapid spread of information, which is easily available through social media platforms such as Twitter. Users of Twitter can send and read content in the form of text or videos using the facilities that Twitter itself offers. Numerous Twitter users have commented on the NCT 127 concert's recent dissolution, which has drawn both supportive and critical remarks. A dataset of 2451 tweets was created by gathering information from Twitter using the keyword "nct" between November 4 and November 6, 2022. The data was subsequently cleaned, yielding a total of 2451 useable data points. Labeling and the Naive Bayes algorithm were then applied to the data. The goal of this study was to count the number of favorable and unfavorable tweets and evaluate how well the Naive Bayes algorithm was applied. According to the trials done, there were 559 favorable remarks and 1,892 negative ones. The accuracy of the evaluation tests was 82.01%. Additionally, the analysis of negative sentiment produced a f1-score of 79.21%, a recall of 68.52%, and precision of 93.84%. Contrarily, the evaluation of positive attitude produced a f1-score of 84.15%, a recall of 95.50%, and a precision of 75.21%. The Naive Bayes method, it may be inferred, can categorize and process with a very consistent accuracy that approaches near-perfect outcomes.