Focusing on the Weakest Link: A Similarity Analysis on Phishing Campaigns Based on the ATT&CK Matrix

Table 2

Literature review on phishing techniques pertaining to the URL.

Focus	Study	Research design	Major findings
URL	Jeeva and Rajsingh (2016)	Experiment done by using an input data set of 1,200 phishing URLs and 200 legitimate URLs.	Analyzed phishing URL to figure significant features to discriminate between legitimate and phishing URLs based on apriori and predictive apriori rule generation algorithm.
	Jain and Gupta (2016)	Dataset of 1,120 phishing URLs and 405 legitimate URLs were used to evaluate the performance of the proposed approach. The URLs were collected between June 2015 and November 2015.	Research focused on fast access time for a real-time environment and high detection rate based on auto updated whitelist of legitimate web sites. The whitelist consists of access of individual users. The performance of the phishing URL detection showed 86.02% of true positive rate while 1.48% of false negative rate.
	Sonowal and Kuppusamy (2020)	Collected 667 phishing URLs from PhishTank, 995 legitimate URLs from Phishload in November 2016.	PhiDMA, multilayer model to detect phishing based on hybrid approach that incorporates 5 layers of whitelist layer, URL feature layer, lexical signature layer, string matching layer, and accessibility comparison layer. PhiDMA achieved accuracy of 92.72% to detect phishing URLs.
	Johnson et al. (2020)	Used ISCX-URL-2016 dataset to train deep learning frameworks.	Compared the performance of the state-of-the-art deep learning framework models and traditional machine learning algorithms.