IPUDCRNN: Integrated Phished URL Detection Using Convolutional Recurrent Neural Network—1D CNN + LSTM

Phishing attacks represent a significant and ongoing threat to online security. To address these attacks, machine learning techniques have proven to be optimal. These methods are not always reliable because phishing URLs can precisely resemble legitimate ones, making them hard to differentiate, and sometimes even some advanced machine learning techniques fail. This study proposes a novel approach that combines techniques of deep learning to enhance phishing detection. Deep learning algorithms are better in identifying patterns in data. We assess the current approach using a dataset that contains phishing URLs and demonstrate that the proposed method achieves improved accuracy of 95.33% in identifying phishing URLs. The proposed approach is also effective in detecting new and previously unknown phishing URLs.