<![CDATA[The rapid growth of information and communication technology has a significant impact on the level of cybercrime. The internet, which was originally used to expedite the exchange of information, is also misused by irresponsible parties. One of the prevalent forms of crime is scams, which are fraudulent activities aimed at gaining unlawful profits by exploiting victims through various tactics. The purpose of this research is to evaluate and compare the performance of three algorithms: Support Vector Machine (SVM), Random Forest, and Naive Bayes in analyzing public sentiment regarding scam cases on social media Twitter. The dataset consists of 9,132 tweets, which undergo preprocessing stages such as cleaning, case folding, and word normalization, leaving 8,879 tweets for analysis. Then, the Synthetic Minority Over-sampling Technique (SMOTE) is applied, with the dataset divided into 80% for training and 20% for testing. The results show that before applying SMOTE, the SVM algorithm achieved the highest accuracy at 82%, followed by Random Forest at 79%, and Naive Bayes at 74%. After applying SMOTE, accuracy significantly increased, with SVM reaching 88%, Random Forest at 84%, and Naive Bayes at 76%. This demonstrates that in sentiment analysis of scam cases, the SVM method achieves higher accuracy than both Random Forest and Naive Bayes.]]>
Copyrights © 2025
