<![CDATA[Phishing is one of the most dangerous attacks in the cyber world. Very few researchers have focused on social media phishing, although SMS phishing can be related to the messaging features available on various social media platforms. This study will utilize PSO and PCA techniques to optimize the performance of RF in social media phishing. This study will compare the performance of PSO and RF with that of PCA and RF. An optimized phishing message detection model was built using NLP, incorporating TF-IDF for feature extraction, PCA and PSO for feature optimization, and Random Forest as a classifier to distinguish phishing messages from normal messages. The RF model optimized by PSO produces nearly balanced metrics: precision (0.9877), recall (0.9728), and F1 (0.9802), all of which are high. The RF model with PCA optimization achieves a slightly lower Accuracy (0.9639) and the lowest Precision (0.9585). Although there were no significant differences in the classification process, PSO and PCA made a real contribution to future research development.]]>
Copyrights © 2025
