With the advancement in technology, the internet has become a platform for wide range of illegal activities ranging from spam advertising to financial fraud. Some of these activities are carried out by embedding malware programs in the URLs. Blacklisting services classify URLs but, the constant creation of newer websites poses a challenge. To overcome this challenge Machine Learning approach is used to classify URLs as malicious or benign. The URL dataset, after addressing the issue of class imbalance is fed into several classification models built using a plethora of classification algorithms. Further, feature selection technique is incorporated to reduce the number of features required for classification and used to rank them based on their importance. Also, rule mining algorithms such as Apriori, FP-Growth and Decision Tree Rules is used to generate IF-THEN rules which helps to establish relationship among the features.
A Machine Learning Approach for Detecting Malicious Websites using URL Features
2019-06-01
3318430 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Detecting Malicious Assembly with Deep Learning
IEEE | 2018
|Detecting Phishing Websites using Data Mining
IEEE | 2018
|Real-world IP and network tracking measurement study of malicious websites with HAZOP
British Library Online Contents | 2017
|