<![CDATA[The phenomenon of insecurity as a psychological symptom is increasingly becoming a topic of discussion on social media. This study aims to analyze public sentiment toward the phenomenon of insecurity as expressed through posts on Platform X. The use of sentiment analysis in the context of insecurity is crucial because the phenomenon is subjective and often undetectable in real life. In this context, sentiment analysis is an effective tool for systematically and objectively exploring user sentiment. Data was collected from Indonesian-language tweets in January 2025 using related keywords such as “insecure”,“minder”, and “overthinking.” After undergoing text preprocessing, the data was classified into three sentiment categories: positive, neutral, and negative. Logistic regression was employed as the classification method, with 10-fold cross-validation used to evaluate model performance. The study’s results show a dominance of negative sentiment at 73.34%, with positive and neutral sentiments accounting for 20.38% and 6.28%, respectively. The model’s average accuracy reached 83.13%, with the best performance in detecting negative sentiment. Wordcloud visualizations revealed a dominance of negatively nuanced words such as “takut”,“rendah” and “sendiri.” These findings underscore the importance of deeper understanding of the psychological dynamics of the digital public. This study also paves the way for data-driven interventions to support mental health literacy in online spaces.]]>
Copyrights © 2025
