RÉSUMÉ
The straightforward demonstration of surfing on the web can turn into an overwhelming task making it almost difficult to get away from hackers and their assaults. Numerous conventional procedures are active to prevent clients from tapping on a malevolent URL, opening loathsome connections, or drawing in with deluding promotions can prompt deadly results. This paper centres on recognizing and classifying Malicious URLs since this is a strong and compelling method to stop assaults. Assailants regularly attempt to transform at least one parts of URL involving them for assaults on client's framework by Drive-by download, phishing, social designing and spam. The created framework will utilize a regulated machine learning approach. Our framework design is isolated into two modules: the initial one is preparing and the second is identification. This framework is carried out as an augmentation for internet browsers to make it user centric.