<![CDATA[This study was motivated by rising public complaints regarding the habit of doomscrolling on social media, particularly TikTok, a practice often linked to negative impact on users mental health. Public opinion on this phenomenon has yet to be systematically mapped, especially given the informal and unstructured nature of TikTok comments. This research aims to analyze sentiment regarding the doomscrolling phenomenon through TikTok comments using a lexicon based approach. Data were collected by retrieving TikTok videos using the search terms “doomscrolling” and “TikTok addiction” via scraping techniques, followed by text preprocessing, automated sentiment labeling using the InSet Lexicon, feature extraction via Term Frequency Inverse Document Frequency (TF-IDF), and classification using the Support Vector Machine (SVM) algorithm. The results indicate that negative sentiment dominates at 70.72%, followed by neutral sentiment 16.57%, and positive sentiment at 12.71%. The classification model achieved an accuracy of 75.76%, a precision of 76.66%, a recall of 75.76%, and an F1-score of 76.07%. These findings suggest that public generally perceives doomscrolling as a troubling habit, and the developed model proves effective for mapping public opinion on the doomscrolling phenomenon on TikTok.]]>
Copyrights © 2026
