Indonesia is one of the active supporters of social networking users. Since
2014, the number of active users has been growing every year. One of the most
used social media is the social media application Twitter with a total of 22.8
billion users in 2019. Of course, many public opinions contain a lot of important
information. Therefore, it is necessary to process information from the database.
The discipline used to study this problem is text mining. The political movement
ahead of the 2024 general election is getting more massive. Several parties and
axes of power determine the number of presidential and vice presidential
candidates. Several political figures have taken advantage of the use of
billboards, including the Chairperson of the Golkar Party, the Chairperson of the
Indonesian Democratic Party of Struggle, the Chairperson of the Democratic
Party, the Chairperson of the Great Indonesia Business Party, and the
Chairperson of the Democratic Party. National Party. A number of names were
mentioned on Twitter, starting with Anies Baswedan, Ganjar Pranowo and
Prabowo Subianto. These names become keywords or keywords when searching
for tweets related to the 2024 presidential election. Sentiment analysis is the
process of understanding and processing textual data automatically to obtain
opinion data, sometimes in the form of opinion statements, based on models
trained on the data. The classifier used is Naive Bayes which has high
classification accuracy and short classification time so as to speed up sentiment
analysis. The 2024 Presidential Candidate Public Opinion Poll uses the Naive
Bayes classification method used in this study to determine the effectiveness and
optimization of this method to find out public opinion about the 2024 Presidential
Candidates on Twitter social media. The results of the 2024 Presidential Election
Case Opinion Analysis study were extracted from Twitter data from January 1,
2020 to June 25, 2023, resulting in 12,000 pieces of data, which then entered the
data cleaning stage and were classified as positive, negative, neutral. Dictionary
and finally begins with the creation of the Naive Bayes model. Opinion scores are
sorted by highest order, starting with positive sentiment, followed by negative
opinion, and lastly neutral opinion, and the accuracy of the Naive Bayes model is
74%.