Machine Learning Algorithms to Detect Suspicious Domain Names in Internet Security

##plugins.themes.academic_pro.article.main##

P. Maragathavalli
B. Tamilarasi
R. Nivetha
S. Anjali

Abstract

Internet security is an advanced protection solution against unknown threats in application based networks.
Attackers can use a server of commands and controls to exploit communication. So malware detection of domain name is
a serious issue in internet security. The signature based detection technique has been widely used as the main method of
detecting malware, but with obfuscation techniques, it has failed to detect modern malware. Blacklisting is the basic
method in detecting the malicious URLs. The other current Heuristic classification method is an update to the Black
Category. In this process the signatures are matched and checked to find the correlation between the new URL and
current malicious URL signature. Although the malignant and benign URLs can be effectively categorized by both
Black-Listing and Heuristic classification. We cannot cope with the emerging methods of attacking. To overcome these
issues, machine learning techniques for Malicious URL Detection is applied and use a set of URLs as training data, and
depending on the statistical properties, learn a prediction function to recognize a URL as malicious or benign. The
machine learning technique is to train a learning-based prediction mechanism, based on data is used for current machinelearning methods may be categorized as Supervised learning Unsupervised learning, Semi-supervised learning, Support
Vector Machine. The Probabilistic Neural Networks in Machine Learning gives good performance than the primitive
technologies like black listing, heuristic approach.

##plugins.themes.academic_pro.article.details##

How to Cite
P. Maragathavalli, B. Tamilarasi, R. Nivetha, & S. Anjali. (2022). Machine Learning Algorithms to Detect Suspicious Domain Names in Internet Security. IIRJET, 5(3). https://doi.org/10.32595/iirjet.org/v5i3.2020.115