A novel robust convolutional neural network for uniform resource locator classification from the view of cyber security

Uniform resource locator (URL)-based cyber-attacks form a major part of security threats in cyberspace. Even though the experience and awareness of the end-users help them protect themselves from these attacks, a software-based solution is necessary for comprehensive protection. To this end, a novel robust URL classification model based on convolutional neural network is proposed in this study. The proposed model classifies given URLs into five classes, namely, (i$$ \mathrm{i} $$) benign$$ \mathrm{benign} $$, (ii$$ \mathrm{ii} $$) defacement$$ \mathrm{defacement} $$, (iii$$ \mathrm{iii} $$) phishing$$ \mathrm{phishing} $$, (iv$$ \mathrm{iv} $$) spam$$ \mathrm{spam} $$, and (v$$ \mathrm{v} $$) malware$$ \mathrm{malware} $$. The proposed model was trained and evaluated on a gold standard URL dataset comprising of 36,707$$ \mathrm{36,707} $$ samples. According to the experimental result, the proposed model obtained an accuracy as high as 98.1%$$ 98.1\% $$ which outperformed the state-of-the-art. Based on the same architecture, we proposed another classifier, a binary classifier that detects malicious URLs without dealing with their types. This binary classifier obtained an accuracy as high as 99.3%$$ 99.3\% $$ which outperformed the state-of-the-art as well. The experimental result demonstrates the feasibility of the proposed solution.