Methods to Identify the Family of Advanced Persistent Threats Based on Deep Neural Network and n-gram of API calls
Bibliographic record
Abstract
Advanced Persistent Threat attacks (APTs) have a set of special features that make them different from other attacks. They are stealth, target-based, and designed by expert teams. There are various methods to detect malwares, but since the APT attacks are complex and well-designed to evade detection with a minimum fingerprint on the target systems, they are difficult to detect. Research on APTs has two challenges: the complexity of the attack and the small number of identified samples. These challenges have some drawbacks to the accuracy of malware classifiers. We address these two challenges in this study by presenting two methods. We used static and dynamic malware analysis as an input to the deep neural network to address the complexity of the attack challenge and n-gram of API calls as an input to four machine learning algorithms to address the small number of available samples. The feature selection phase of these two methods employs the TF-IDF measure to identify valuable terms in the malware analysis report. The proposed hybrid approach shows better results for the dataset we used for this research compared to the other baseline methods. In this research, we collected an APT dataset with a clear validation method which will benefit our future research.
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 | 0.002 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 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".