Phishing Detection and Prevention Using Natural Language Processing (NLP)

Kokkalakonda, Manideep (2025) Phishing Detection and Prevention Using Natural Language Processing (NLP). Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (2MB) | Preview

Abstract

Phishing remains one of the most prevalent cybersecurity threats, with email serving as the primary attack vector. Traditional detection methods—such as blacklists, keyword matching, and rule-based systems—struggle to detect evolving attacks that mimic legitimate communication and exploit subtle social engineering tactics. This research presents a hybrid phishing email detection model that integrates BERT-based semantic embeddings with structural and stylistic features, processed through a BiLSTM architecture to capture both contextual meaning and sequential dependencies.

The study compiles a unified dataset from multiple publicly available phishing and legitimate email sources, totaling 164,971 emails (85,781 phishing and 79,190 legitimate), and applies comprehensive NLP preprocessing, and extracts a 782-dimensional feature vector combining 768 BERT embedding dimensions with 14 handcrafted features (e.g., special character ratio, URL count, money-related terms).

The proposed hybrid model is evaluated against traditional TF-IDF-based machine learning baselines, including Logistic Regression, Random Forest, and Gradient Boosting. Experimental results demonstrate high accuracy (95.7%), strong recall for phishing detection (96.35%), and excellent ROC-AUC performance (0.9916).

Explainable AI techniques (SHAP, LIME) provide feature-level insights, revealing that urgency keywords, monetary terms, and special character usage are key phishing indicators. The model offers a scalable, interpretable, and enterprise-ready framework for real-time phishing detection.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Aleburu, Joel UNSPECIFIED
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing Q Science > QA Mathematics > Computer software > Computer Security T Technology > T Technology (General) > Information Technology > Computer software > Computer Security H Social Sciences > HV Social pathology. Social and public welfare > Criminology > Crimes and Offences > Cyber Crime
Divisions:	School of Computing > Master of Science in Cyber Security
Depositing User:	Ciara O'Brien
Date Deposited:	15 Jun 2026 14:24
Last Modified:	15 Jun 2026 14:24
URI:	https://norma.ncirl.ie/id/eprint/9357

Actions (login required)

View Item