Machine Learning for SPAM Detection - Review Article - Volume 6, Issue 1

Aug 2, 2026

Upload by Noah

Email and messaging systems underpin communication across industries, yet spam—often termed junk or unwanted mail—continues to grow and harms users by consuming time and computing resources while risking important data. Providers face major challenges in spam identification and filtration, making spam filtering a core and widely used protection approach. The survey reviews machine learning and deep learning methods for spam filtering, including Naive Bayes, decision trees, neural networks, and random forests, and compares them via accuracy, precision, recall, and related measures.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages13

Reading Time33 min

Introduction

What Is Spam?

What Is Ham?

SPAM Filtration Techniques

The Common Spam Filtering Technique

Filtering of Spam on the Client Side

(Further techniques)

Asian Journal of Advances in Research Volume 6, Issue 1, Page 167-179, 2023; Article no.AJOAIR.2465 Machine Learning for SPAM Detection Phani Teja Nallamothu a* and Mohd Shais Khan b a Strava, United States. b Osmania University, Hyderabad, Telangana, India. Authors’ contributions This work was carried out in collaboration between both authors. Both authors read and approved the final manuscript. Review Article Received: 09/01/2023 Accepted: 15/03/2023 Published: 17/03/2023 ABSTRACT In practically every industry today, from business to education, emails/messages are used. Ham and spam are the two subcategories of emails/messages. Email or message spam, often known as junk email or unwelcome email, is a kind of message that can be used to hurt any user by sapping their time and computing resources and stealing important data. Spam messages volume is rising quickly day by day. Today's email and IoT service providers face huge and massive challenges with spam identification and filtration. Spam filtering is one of the most important and well-known methods among all the methods created for identifying and preventing spam. This has been accomplished using a number of machine learning and deep learning techniques, including Naive Bayes, decision trees, neural networks, and random forests. By categorizing them into useful groups, this study surveys the machine learning methods used for spam filtering. Based on accuracy, precision, recall, etc. , a thorough comparison of different methods is also made. Keywords: Spam; ham; machine learning; supervised machine learning. 1. INTRODUCTION These days, short message service is a very popular method of communication. This system has millions of users linked to it because of its quick responses, accessibility, and affordable costs. There are two different types of SMS [1] . The first is spam, which counts the number of unsolicited commercial messages a user has received. With these notifications, the user encounters a number of issues, including a slow device and storage concerns [2] . Further, deleting spam from memory takes a long time. Various techniques, such as blacklist, naive bluesman, and keyword matching algorithms, are utilized to identify this spam issue [3-5] . Spam communications have a negative effect on text and email messages today and annoy SMS users. Cybercriminals and various advertising agencies utilize these kinds of spam [6] . *[Corresponding author: Email: Phani.teja89@gmail.com](Corresponding author: Email: Phani.teja89@gmail.com); Asian J. Adv. Res., vol. 6, Issue 1, Page 167-179, 2023 Machine Learning Method Supervide Methods UnSupervised Methods Deep Learning Methods Bert Technique Spam detection by using DL Application Fig. 1. Machine learning techniques The fundamental problem with spam is that thereis no longer any privacy; when someone responds to these SMS messages, privacy is violated. Because it uses a single click to assault the privacy bridge [7,8] . The easiest tool for a cyber-attack is a mobile phone. Research has shown that more than 200 million mobile users receive spam SMS in a single day, which is insufficient [9-11] . 1.1 What Is Spam? Unwanted and unpleasant text messages in the form of spam are those that we repeatedly getviaa transmission channel. Spam messages have an impact on a device's performance, power, and storage system. In short, spam has proven to be the most unpleasant aspect of personal communication [12,13] . 1.2 What Is Ham? Ham refers to messages that we receive from end devices that are not spam and are on a good list of requested and wanted messages. About 2001, Spam Bayes first used the term "ham,"which is currently recognized to mean "e-mailand messages that are commonly appreciated and aren't deemed spam [14,15] . Its utilization is especially normal among antispam software developers, and not broadly known somewhere else; as a general rule, it is most likely

FAQs

Why is spam detection important for email and IoT service providers?

Spam messages consume user time and computing resources, and can steal important data. Growing spam volume creates major identification and filtration challenges, especially for email and IoT providers.

What machine learning methods are surveyed for spam filtering?

The survey covers machine learning and deep learning approaches such as Naive Bayes, decision trees, neural networks, and random forests, along with deep learning techniques including BERT.

How does the paper evaluate and compare spam filtering methods?

Methods are compared using common metrics such as accuracy, precision, and recall to assess performance across different approaches.

Aug 2, 2026

Upload by Noah