Phishing Emails Classification Research Trends: Datasets, Features and Methods

Abstract

Phishing is a way for getting sensitive data of someone. Getting the usernames, passwords, and credit card numbers through deceptive media created by the attackers. Phishing email detection and classification is one of the areas that needed to be improved and researched, because the number phishing attack escalated quickly every year and will have a high impact of on the cyber security area. Therefore, this paper aims to provide a comprehensive review on the application of phishing email classification. The literatures of this study are collected articles which the keyword of “Phishing Email Classification”, “Phishing Email Attack”, and “Phishing Email Detection”. This paper targets the literature on five main aspects: (1) phishing email classification application areas and approaches, (2) features analysis, (3) dataset analysis, (4) performance and result measure analysis, (5) and future work. This comprehensive study presents analysis of the phishing emails classification domain by focusing on the dataset, techniques, and features being used on the researches and experiments. The articles published in 2015-2020 were used as the review objects. The literature survey consists a total of 23 papers on the phishing email classification paper; 3 papers in 2020, 8 papers in 2019, 2 papers in 2018, 5 papers in 2017, 3 papers in 2016, and 2 papers in 2015. The paper is organized as follows: Section 1 give briefly explanation all about Phishing. Section 2 presents the analysis of previous research method, area, dataset, preprocessing, approaches and evaluations. Section 3 presents the critical analysis, strength and limitation, and open issues. Section 4 discusses the future research challenges. Section 5 concludes the paper.