The use of artificial intelligence (AI) to detect phishing emails is primarily dependent on large-scale centralized datasets, which has opened it up to a myriad of privacy, trust, and legal issues. Moreover, organizations have been loath to share emails, given the risk of leaking commercially sensitive information. Consequently, it has been difficult to obtain sufficient emails to train a global AI model efficiently. Accordingly, privacy-preserving distributed and collaborative machine learning, particularly federated learning (FL), is a desideratum. As it is already prevalent in the healthcare sector, questions remain regarding the effectiveness and efficacy of FL-based phishing detection within the context of multi-organization collaborations. To the best of our knowledge, the work herein was the first to investigate the use of FL in phishing email detection. This study focused on building upon a deep neural network model, particularly recurrent convolutional neural network (RNN) and bidirectional encoder representations from transformers (BERT), for phishing email detection. We analyzed the FL-entangled learning performance in various settings, including (i) a balanced and asymmetrical data distribution among organizations and (ii) scalability. Our results corroborated the comparable performance statistics of FL in phishing email detection to centralized learning for balanced datasets and low organizational counts. Moreover, we observed a variation in performance when increasing the organizational counts. For a fixed total email dataset, the global RNN-based model had a 1.8% accuracy decrease when the organizational counts were increased from 2 to 10. In contrast, BERT accuracy increased by 0.6% when increasing organizational counts from 2 to 5. However, if we increased the overall email dataset by introducing new organizations in the FL framework, the organizational level performance improved by achieving a faster convergence speed. In addition, FL suffered in its overall global model performance due to highly unstable outputs if the email dataset distribution was highly asymmetric.

Download full-text PDF

Source
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC10181597PMC
http://dx.doi.org/10.3390/s23094346DOI Listing

Publication Analysis

Top Keywords

phishing email
16
email detection
16
organizational counts
16
email dataset
12
federated learning
8
global model
8
neural network
8
increasing organizational
8
counts increased
8
email
7

Similar Publications

Life has become more comfortable in the era of advanced technology in this cutthroat competitive world. However, there are also emerging harmful technologies that pose a threat. Without a doubt, phishing is one of the rising concerns that leads to stealing vital information such as passwords, security codes, and personal data from any target node through communication hijacking techniques.

View Article and Find Full Text PDF

Objectives: Difficulties with deception detection may leave older adults especially vulnerable to fraud. Interoception, that is, the awareness of one's bodily signals, has been shown to influence deception detection, but this relationship has not been examined in aging yet. The present study investigated effects of interoceptive accuracy on 2 forms of deception detection: detecting interpersonal lies in videos and identifying text-based deception in phishing emails.

View Article and Find Full Text PDF

With technological advancements, financial exploitation tactics have expanded into the online realm. Older adults may be particularly susceptible to online scams due to age- and Alzheimer's disease-related changes in cognition. In this study, 182 adults ranging from 18 to 90 years underwent cognitive assessment, genotyping for apolipoprotein E e4 (APOE4), and completed the lab-based Short Phishing Email Suspicion Test (S-PEST) as well as the real-life PHishing Internet Task (PHIT).

View Article and Find Full Text PDF

Unsolicited scam invitations from predatory publications and fraudulent conferences: Radiology-in-training experience.

Clin Imaging

September 2024

Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins Medical Institution, Baltimore, MD, United States of America.

Purpose: Radiology faculty across various specialties have been reported to receive an average of 20.7 invitations to submit manuscripts to bogus journals and 4.1 invitations to speak at unsuitable events over a two-week span.

View Article and Find Full Text PDF

This study investigated the roles of phishing knowledge, cue utilization, and decision styles in contributing to phishing email detection. Participants (N = 145) completed an online email sorting task, and measures of phishing knowledge, email decision styles, cue utilization, and email security awareness. Cue utilization was the only factor that uniquely predicted the capacity to discriminate phishing from genuine emails.

View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!