Leveraging Human Knowledge in Large Language Model for Obfuscation-Resisted Phishing URL Detection
Bibliographic record
Abstract
Phishing is a significant and growing threat on the Internet. There is a growing trend of using deep learning techniques for phishing URL detection due to their capability to recognize a wide range of patterns. However, URL obfuscations have been a popular approach for adversaries to escape the detection mechanism. Existing deep learning solutions are limited to the patterns presented in the training set and suffer from robustness issues. Unlike traditional machine learning solutions, we present the first novel approach to phishing URL detection based on pre-trained large-scale language models (LLMs). Leveraging the general human knowledge embedded in the language model, our solution can accurately detect obfuscated phishing URLs without explicitly training them. We are the first to simulate real-world scenarios of obfuscated phishing URLs in testing procedures, using two prevalent techniques: Domain Obfuscation and Redirects. Our model achieves over 92% and 83% in accuracy in two obfuscation test benchmarks, significantly outperforming baseline models. This study offers a robust and adaptable solution to evolving phishing threats.
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 | 0.001 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 0.002 |
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".