AI-Enabled Phishing Links Detection Using Machine Learning Models
Bibliographic record
Abstract
Phishing attacks have grown into a widespread threat to cybersecurity, exploiting the rapid growth of the internet and digital resources. These attacks typically mislead victims into disclosing private information such as credit card numbers, passwords, and login credentials. Because they rely on static patterns, most rule-based detection techniques struggle to identify sophisticated phishing attempts. In this study, we propose an AI-enabled phishing detection system using various machine learning models, including Logistic Regression, Random Forest, Gradient Boosting, and Support Vector Machine, along with a non-machine learning heuristic rule-based approach. These models are trained on a comprehensive dataset of over 11,000 URLs. To enhance transparency and interpretability, we incorporate Local Interpretable Model-Agnostic Explanations (LIME), which highlights the most influential features in real-time classification. Model performance is evaluated using metrics such as accuracy, precision, recall, and F1-score. Our results show that ensemble-based models particularly Random Forest and Gradient Boosting outperform traditional classifiers in effectively distinguishing phishing URLs from legitimate ones. The study also provides recommendations for future research and emphasizes the impact of feature engineering and data preprocessing on improving detection accuracy.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".