Analysis of Public Sentiment Towards Retired Military Officers' Pressure to Impeach the Vice President Through X Using Decision Tree

The advancement of information technology has strengthened the analysis of social media data, particularly in understanding public opinion on national political issues. This study examines public sentiment on the X (Twitter) platform regarding the issues of Gibran’s Impeachment and the Urging by Retired Military Officers using the Decision Tree CART algorithm. Data were collected through a crawling process, resulting in 1,020 tweets for the Impeachment issue and 89 tweets for the Urging issue. After preprocessing, the dataset was labeled using a Lexicon-Based method that classifies text into positive, negative, and neutral sentiments. The evaluation results show that for the Impeachment issue, the model achieved an accuracy of 97.05%–99.51%, with the highest performance found in the Neutral class (F1-Score 98.46%). For the Urging issue, the model obtained an overall accuracy of 88.89%, with the highest performance also in the Neutral class (F1-Score 94.12%). Model performance decreased in the Positive and Negative classes due to data imbalance. Overall, the findings indicate that Decision Tree CART is effective for sentiment classification on small to medium datasets and reveal that public sentiment toward both issues is predominantly Neutral.