In spite of various research endeavors, phishing assaults stay common and exceedingly successful in attracting clueless clients to uncover delicate data, including account details and government managed savings numbers. Misfortunes due to phishing are developing consistently. A solitary methodology isn't effective for distinguishing a wide range of phishing assaults. So we propose a hybrid approach to deal with the classification of URLs as phishing or real. The investigation aftereffects of our proposed methodology, in view of a dataset gathered from phishing and legitimate URLs, have demonstrated that PhishAlert framework can successfully counteract phishing assaults and can thus ensure system security.