Improving Malicious PDF Detection with a Robust Stacking Ensemble Approach
Bibliographic record
Abstract
In recent years, the increasing prevalence of malicious PDF files has become a major cybersecurity concern. Despite advances in machine learning-based detection methods, cyber attackers continue to develop novel techniques for evading detection, necessitating the development of more robust and effective models. In this paper, we propose a stacking ensemble model that is used as part of a framework for robustness against feature-specific attacks. Additionally, we address the limitations of the PDFMal-2022 dataset by enhancing the feature extraction module and resolving issues related to flawed values and mismatched mapping, creating an improved dataset with additional features, the enhanced PDFMal-2022. We evaluate our model’s performance on the widely used Contagio dataset and compare it to three state-of-the-art approaches. Moreover, we provide a benchmark performance of the proposed model on the enhanced PDFMal-2022 dataset, validating its effectiveness in a more challenging setting. And we demonstrate our proposed framework’s performance and robustness against malicious PDFs from the QAKBOT and ICEDID malware campaigns. Our proposed stacking ensemble model and the enhanced PDFMal-2022 dataset contribute to the field of malicious PDF detection, providing a valuable asset in combating PDF-based cyber 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 | 0.002 | 0.004 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.003 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| 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".