This research examines opinion sentiment regarding (1) voters in the 2024 election using data analysis from the social media Twitter. (2) Using a text mining and classification approach, (1) this research extracts valuable information from tweets containing keywords related to the 2024 election. The data collection process is carried out using scraping techniques, where tweets are collected within a certain period of time to ensure complete representation. After the data is collected, (2) preprocessing is carried out to clean and prepare the text, which includes steps such as tokenaize, stopword and labeling. (1) Sentiment analysis is then used to categorize tweets into positive, negative or neutral sentiment. (2) The K-Means algorithm is used to collect opinion data to help identify patterns and trends in public perception of political candidates and issues. (1) Analysis results shows that there is a significant distribution of opinions between different candidates and issues, thus revealing the complex dynamics of public opinion. (2)These results provide policymakers, political candidates, and researchers with an in-depth understanding of how public opinion is formed and how it can be influenced during election campaigns. Additionally, this research highlights the great potential of applying text mining technologies and algorithms