| dc.description.abstract | Phishing attack is a chronic cybersecurity threat, particularly in popular platforms
like Gmail, where traditional rule-based systems often struggle to detect evolving
phishing tactics. This review examines effectiveness of the approaches of AI/ML-based
Gmail plugins focusing on supervised learning and NLP techniques. These technologies
increase the accuracy and flexible of phishing email detection while fixing the issues with
traditional methods. The literature review was conducted by considering publications
between 2015 and 2024 and the materials retrieved from trusted cybersecurity websites
and other scholarly sources. The research points out the strengths and weaknesses
by contrasting them with alternative approaches, so the use of AI/ML-based plugins
enhances e-mail security, which is an advantage over rule-based, static systems. This
research highlights that the Supervised Learning and NLP improve the detection of
phishing emails by focusing on patterns of the email content where traditional methods
are failed to identify. Future research is required to carryout overcome the constraints
of real-time automatic responses for more flexibility. This review recommends for
future developments with the combination of deep learning and privacy-preserving
federated learning to improve real-time response capabilities. The circle of usage could
be increased by developing better user interfaces, user alert systems, and reporting
tools. These would make email security solutions with more agile and efficient, while
improving phishing detection, and also improving the security of the email ecosystem. | en_US |
