Rotation Forest model modification within the email spam classification

А.О Шанін

doi:10.30748/soi.2021.164.12

Abstract

Increased use of email in daily transactions for many businesses or general communication due to its cost-effectiveness has made emails vulnerable to attacks, including spam. Spam emails are unsolicited messages that are very similar to each other and sent to multiple recipients randomly. This study analyzes the Rotation Forest model and modifies it for spam classification problem. Also, the aim of this study is to create a better classifier. To improve classifier stability, the experiments were carried out on Enron spam, Ling spam, and SpamAssasin datasets and evaluated for accuracy, f-measure, precision, and recall.

Highlights

– True Positive (TP) – the number of emails with spam is correctly defined as spam;
The formula is defined as follows: Recall = TP / (TP + False Negative (FN)). – Precision defined as a fraction of correctly detected spam messages relative to all messages which detected as spam
The Rotation Forest method was chosen for improvement, experiments with which showed that the use of another basic algorithm can significantly improve the efficiency of this metamodel

Summary

Introduction

Email is a extremely fast and cost-effective mean of transmitting information from anywhere in the world, which can be used from personal computers, smartphones and other electronic gadgets of the latest generation [1]. In 2021, the number of email accounts worldwide was estimated at 4 billion, which is more than half of the world's population [2]. The growing popularity and use of emails for transactions has led to an increase in spam worldwide. Spam emails are unsolicited messages sent by email to several recipients who did not wish to receive these messages. Since email messages are the main means of sending harmful information, including viruses and phishing attacks, the number of spam messages is growing rapidly and is one of the most serious threats to email users [7,8]

Objectives

Results

Conclusion