Efficient Feature Selection Approach for Detection of Phishing URL of COVID-19 Era

ABSTRACT

Cybercrime is a growing concern, particularly in this COVID-19 era. The COVID-19 outbreak has shown the significant impact potential of such crises on our daily lives worldwide. Phishing is a social engineering crime that can cause financial and reputational damages such as data loss, personal identity theft, money loss, financial account credential theft, etc., to people and organizations. In the recent outbreak of the COVID-19 pandemic, many companies and organizations have changed their working conditions, moved to an online environment workspace, and implemented the Work From Home (WFH) business model that increases the phishing attacks vectors and risk of breaching internal data. In this paper, we have extracted nine efficient features from the URLs and applied seven different Machine Learning algorithms to recognize phishing URLs. Machine learning algorithms are often used to detect phishing attacks more accurately before affecting users. The obtained result concludes that the Random Forest model provides the best and highest accuracy of 95.2%. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.