<![CDATA[The rapid expansion of Islamic preaching in the digital sphere, particularly through YouTube, calls for a deeper understanding of communication ethics as reflected in user responses. This study analyzes the sentiments expressed in comments on Islamic preaching videos to identify patterns of digital ethics within online communities. The research employs a Support Vector Machine (SVM) classification model with TF-IDF feature representation. Data were collected from YouTube comments and processed through several preprocessing stages, including text cleaning, case normalization, tokenization, stopword removal, and stemming, before being manually labeled into three sentiment categories: positive, negative, and neutral. Testing on 22 data samples shows that the SVM model achieved an accuracy of 77.27%, with the highest performance observed in the neutral category. Misclassification in the positive and negative categories was mainly influenced by data imbalance and linguistic variations commonly found in religious discourse. These findings indicate that SVM combined with TF-IDF is reasonably effective for sentiment analysis in the context of digital Islamic preaching; however, improvements in data balance and the incorporation of contextual features are necessary to enhance classification performance. Overall, this study provides an initial insight into audience response patterns toward digital Islamic preaching and contributes to the development of digital ethics research in Islamic communication studies.]]>
