Article Preview
Top1. Introduction
Social media platforms have recently become a primary source of news and information (Wang et al., 2022; Li, Zhou & Huang, 2021). However, the ease of sharing and spreading information on social media has led to an increase in the spread of fake news, which can significantly impact public opinion, politics, and society (Apuke & Omar, 2021; Wang et al., 2023; Naeem, Bhatti & Khan, 2021; Meng, Xiao & Wang, 2022). Detecting fake news in email is challenging, as the content can be diverse, misleading, and constantly evolving (Zhang & Ghorbani, 2020; Huang, 2020; Li et al., 2022). Machine learning algorithms have shown promising results in detecting fake news on email platforms (Najadat, Tawalbeh & Awawdeh, 2022) and other domains (Zhou et al., 2022; Zheng et al., 2021; Zheng & Yin, 2022; Chen, Chen & Lu, 2023). This paper presents a machine-learning technique for detecting fake email news using the Support Vector Machine with Radial Basis Function (SVM-RBF) and K-Nearest Neighbor (KNN) algorithms.
Fake news has become a significant problem on email and other platforms, with the potential to cause harm to individuals and society (Zhang & Ghorbani, 2020). The spread of fake news can lead to confusion, misinterpretation, and damage to reputation. Fake news can also be used to manipulate public opinion and affect the outcome of political processes (Gradon, 2020; Guo & Zhong, 2022). Hence, detecting and preventing phony news spread is crucial to maintaining a healthy and informed society. Several techniques have been proposed to detect fake news in spam emails, including manual fact-checking, crowdsourcing, and machine learning (Chen, Lai & Lian, 2022). Manual fact-checking and crowdsourcing techniques can be time-consuming, costly, and error-prone, especially when dealing with a large volume of data (Wu et al., 2022).
On the other hand, machine learning algorithms have shown great potential in detecting fake news in spam emails (Zhang & Ghorbani, 2020) and some other fields like financial transactions (Wu et al., 2023; Li & Sun, 2020), construction (Qin et al., 2022), summarization (Deng et al., 2023). Methods like big data analytics have become a powerful tools in analyzing large datasets in different domains such as residents happiness (Li et al., 2023), risk of SME shutdown (Xie et al., 2023), visual chirality cue (Tan et al., 2023), networked control systems (Zhong et al., 2022) etc. Several prior efforts have been made to address the issue of counterfeit news detection, including manual fact-checking, third-party fact-checking services, and machine-learning techniques (Babaei, 2021). However, manual fact-checking is time-consuming and may not scale well.
Third-party fact-checking services can be effective, but they may not cover all the news items posted on social media platforms (Oeldorf-Hirsch et al., 2020; Ardevol-Abreu, Dwlponti & Rodriguez-Wanguemert, 2020). Machine learning techniques, including Natural Language Processing (NLP) and Deep Learning, have been utilized to detect fake news spam emails in a cloud (Althubiti, Alenezi & Mansour, 2022; Zhang et al., 2023) and overhead catenary (Madani, Motameni & Mohamadi, 2022; Zong, Wang & ZhiboWan, 2022; Efanov et al., 2016). While these approaches have shown promise, they may not always be practical, mainly when dealing with news items not in English (Zong & Wan, 2022).