Detection and Classification of Legitimate and Spam Emails using K-Nearest Neighbor Augmented with Quadratic Sieve Algorithm

Abstract

internet as it endangers financial institutions and poses a threat to individual users. Various techniques have been proposed by different studies to prevent spam in emails; however, classification and filtering technique using machine intelligence methods are the most efficient among the several methods. This study employed a K-Nearest Neighbor (KNN) augmented with the Quadratic Sieve algorithm to detect and classify legitimate emails and spam. The sieve algorithm revealed all the prime numbers for all the dataset used, starting from the input dataset to reduce the errors that may cause an imbalance in the classification. The result from this study shows that implementation of KNN augmented with Quadratic Sieve algorithm detects and properly classifies legitimate e-mail as well as spam much better.