Evaluation of Classification Algorithms for Effective Spam Email Detection using Spam Email Dataset
Kalyani Shivaji Ubale, Kamini Shirsath · 2025
This Abstract- The rapidly evolving digital landscape has resulted in a surge of spam emails, posing significant cybersecurity threats through methods like social engineering, malware distribution, and phishing. Traditional spam filtering techniques have proven insufficient in addressing the complexity of modern spam tactics. This paper presents a detailed evaluation of various classification algorithms on publicly available spam email datasets. By analyzing these datasets, we offer insights into the most effective data mining approaches for spam identification and prevention. The study involves a comparative analysis of classifiers such as Naive Bayes, K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Logistic Regression, Decision Tree, and Random Forest. Our work not only examines the classifiers’ performance but also highlights the datasets’ characteristics, exploring factors like spam email percentage, the number of occurrences, feature types, and data attributes. The novelty of this research lies in its comprehensive approach to understanding spam email patterns, its emphasis on the dynamic nature of spam, and its use of diverse datasets to validate the effectiveness of different machine learning algorithms.