The New Phishing Frontier: Predictive Modeling of Malicious Click Trajectories Originating from Social Networking Sites

DOI

10.2991/978-94-6239-723-1_9How to use a DOI?

Keywords

Phishing Attacks; Cybersecurity; Social Engineering; Psychological Elements; Sociotechnical Elements; Human-Centric Security; Human Vulnerabilities; Risk Mitigation; Economic Repercussions; Reputational Harm

Abstract

In this paper, we present and test predictive models that are specifically targeted at malicious clicks in order to solve the problem of cybercrime, which has become a widespread problem among social media users. Using a number of individual features like the behavior of the user, device used, social media platform, reputation of the website, and others, we developed a predictive model for Click_Leads_To_Malicious_Site (binary). Our model exhibits high predictive capability in all data sets (TRAIN, VALIDATE, and TEST). It demonstrates very high discriminative power with some excellent performance measures such as Area Under the Curve (AUC) of 0.9623 on the VALIDATE data set and 0.9462 on the TEST data set. The model performs very well on the VALIDATE data set with respect to the default cutoff point (0.5) at 0.916. Another excellent metric is the F1 score of 0.807 on the VALIDATE data set, which demonstrates excellent precision and recall. The model performance is proven by Lift and Captured Response Percentage returning. The model has identified a Cumulative Lift of 4.26 in the VALIDATE partition at 10% depth (top 10% of the data based on the predicted probability) that means to detect 4.26 times more bad cases compared to the randomly selected one. The higher value of lift means the model is capable of detecting priority high-risk transactions. Considering the fact that the focus here is on avoiding risk and creating pain for the participants, this model could help improve the security of customers in the online environment in real-time by assessing the risk of malicious clicking on social accounts.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Mirza Samiulla Beg
PY  - 2026
DA  - 2026/07/14
TI  - The New Phishing Frontier: Predictive Modeling of Malicious Click Trajectories Originating from Social Networking Sites
BT  - Proceedings of the International Conference on Responsible, Risk-aware, and Regulated AI (RRRAI 2026)
PB  - Atlantis Press
SP  - 92
EP  - 106
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-723-1_9
DO  - 10.2991/978-94-6239-723-1_9
ID  - Beg2026
ER  -