<![CDATA[The suspicion of radioactive contamination in the Cikande industrial area has prompted a strong public reaction, as shown by the many comments on TikTok. This research aims to understand how people feel about this issue and to measure how well the Support Vector Machine (SVM) algorithm classifies opinions. The data used comes from 3,160 comments collected via web scraping, then processed through several steps, including cleaning text, normalizing, separating words, deleting common words, and summarizing words, before using the TF-IDF method for representation. Comments were then labelled using a lexicon-based method, which showed that 70.79% were negative and 29.21% were positive. Modelling was carried out using SVM on training and test data in an 80:20 ratio. The results show that the model achieved an accuracy of 89% and recognized both sentiment types well. In general, negative comments expressed greater concern about health and environmental impacts, and a lack of confidence in how waste is managed, while positive comments emphasized the importance of scientific verification and official monitoring. These findings indicate the need for clearer, more consistent, and data-based communication about risks to reduce public concerns and increase public trust.]]>
Copyrights © 2026
