Résumé:
Spam emails are an ever-increasing issue in the digital communication world, posing
significant threats globally, ranging from privacy breaches to financial losses and the
spread of malicious software. The sheer volume and evolving sophistication of spam
necessitate robust detection systems to protect users and organizations. In this study,
we introduce an advanced approach to spam email detection using the Improved
Fuzzy K-Nearest Neighbors (IFKNN) algorithm. The primary objective is to enhance
the accuracy and robustness of email classification systems in distinguishing spam
from legitimate messages. Our method integrates fuzzy logic principles to account
for the uncertainty and vagueness inherent in spam classification. We thoroughly
evaluate the performance of IFKNN by comparing it with traditional algorithms
such as KNN and Support Vector Machines (SVM) on real-world spam datasets.
The experimental results demonstrate that IFKNN significantly outperforms existing methods in terms of accuracy, precision, recall, and F1-score, thus showcasing its
potential in advancing cybersecurity measures
