<![CDATA[This study analyzes public perceptions of online gambling through data obtained from the social media platform X (Twitter). A total of 5,846 tweets were collected using the Tweet-harvest tool, which leverages authentication tokens to efficiently gather large datasets. After the pre-processing stage, 4,782 relevant tweets were analyzed using the VADER (Valence Aware Dictionary and Sentiment Reasoner) method for sentiment labeling, categorizing tweets as either positive sentiment or negative sentiment. The sentiment labeling results revealed that 4,151 tweets (86.9%) contained negative sentiment, while 631 tweets (13.2%) contained positive sentiment. The labeled data were trained using a Support Vector Machine (SVM) model with a linear kernel, achieving an accuracy of 97%, with a negative sentiment precision of 97%, recall of 100%, and a positive sentiment precision of 99% with recall of 79%. The confusion matrix analysis demonstrated that the model correctly predicted 1,253 out of 1,255 negative sentiment tweets, while for positive sentiment, it accurately predicted 142 out of 180 tweets. These findings highlight the dominance of negative sentiment in public perceptions of online gambling, associated with potential social impacts such as addiction and economic losses. This research provides valuable insights and data-driven recommendations for policymakers to develop more responsive strategies to address the challenges of online gambling in Indonesia.]]>
