Revisão sistemática da literatura em artigos sobre detecção de phishing: comparação entre técnicas tradicionais e baseadas em inteligência artificial

Abstract

This study investigates phishing detection in the context of cybersecurity, focusing on comparing traditional techniques with methods based on Artificial Intelligence (AI). The primary objective is to conduct a systematic literature review to analyze and compare different detection approaches, with special emphasis on identifying attacks in social media environments. The research highlights the challenges and limitations of traditional techniques such as blacklists and heuristic analysis, while exploring the potential of modern approaches, including machine learning, deep learning, and natural language processing. The methodology used applies the PICOC method to structure the systematic review and address the established research questions. Preliminary results indicate that AI-based techniques offer higher accuracy and adaptability in the face of sophisticated attacks, although they face challenges such as the need for large data volumes and computational complexity. This study contributes to the advancement of cybersecurity strategies by proposing practical recommendations for mitigating phishing attacks in dynamic environments like social media.