A Survey on Phishing URL Detection

Phishing and malicious URLs have become an extensive concern for cybersecurity focusing on individuals, businesses, and government entities. Machine learning has therefore been identified as a robust solution for identifying automatically and precisely these types of threats. This literature review summarizes results from five recent papers that investigate several machine learning and deep learning methods, such as CNNs, LSTMs, Random Forests, Gradient Boosting, k-NN, Logistic Regression, and SVMs. They highlight feature engineering, binary visualization, real-time detection systems, and hybrid schemes in achieving high detection accuracy. They also accentuate the disadvantage of conventional blacklisting techniques and underpin the use of URL lexical features, domain-based heuristics, and image analysis-based techniques for their effectiveness in defeating dynamic phishing tactics. The collective findings attempt to give an overall picture of the present state of affairs in malicious URL detection research and influence future developments within this vital area.