Comparative Evaluation for High Intelligent Performance Adaptive Model for Spam Phishing Detection

Modern day technology, daily seeks to better data processing activities through features such as improved speed, better functionality, higher mobility, portability and improved data access – all of which is extended via smart computing. The widespread use of smartphone has led to an exponential growth in the volumes of emails, alongside great success in phishing attacks carried out more effective via spam inbox mails to unsuspecting users – soliciting for funds. Many mail apps today, offers automatic filters as a set of rules to help better organize and dispose (as spam, if necessary) incoming mails based through the checking of certain keywords detected in a message’s header or body. Achieving such programming filter feature is quite mundane and also inefficient, as spams often evade such filters, slipping into inbox again and again. The study seeks to provide an intelligent adaptive mail support that learns user’s preference via an evolutionary unsupervised model(s) as a computational alternative that adapts the data locality feat as well as compares convergence results yielded by the unsupervised hybrid classifiers. It achieves such feats by building local decision heuristics into their classification processes so that such spam filter(s) are embedded with a design that allows for email genres.