Year: 2024
Type: article
Abstract: Phishing, a prevalent and dangerous form of cybercrime, has been a significant threat since it was first identified in 1996.It involves deceptive emails and fraudulent websites designed to steal sensitive information.Despite ongoing efforts to prevent and detect phishing, no comprehensive solution has been established.This study introduces an automated phishing detection framework using machine learning algorithms.The aim is to effectively distinguish phishing from legitimate software while maintaining a low false positive rate.The research focuses on static phishing detection for Windows operating systems, utilizing a dataset of over 11,000 URLs from phishing and legitimate websites.Various machine learning models, including Decision Tree (DT), Logistic Regression (LR), Random Forest (RF), Naive Bayes (NB), Gradient Boosting Classifier (GBM), K-Nearest Neighbor (KNN), and Support Vector Classifier (SVC), are applied.Additionally, a hybrid LSD model, which combines LR, SVC, and DT with soft and hard voting mechanisms, is proposed.The LSD model employs canopy feature selection, cross-fold validation, and grid search for hyperparameter optimization.Evaluation metrics, such as accuracy, precision, recall, F1-score, and specificity, are used to assess model performance.Results demonstrate that the LSD model outperforms other models in terms of accuracy and efficiency.While the proposed system effectively detects phishing, it cannot automatically adapt to evolving phishing techniques.This limitation can be addressed by retraining the model with updated data.Although the study focuses on static analysis, integrating dynamic analysis in future research could further enhance detection capabilities.Overall, the framework provides an effective solution for phishing detection with room for future improvements.
Language: English
Topic: Spam and Phishing Detection
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4405912149