Enhancing GNN Explanations for Malware Detection with Dual Subgraph Matching
Bibliographic record
Abstract
The increasing sophistication of malware has challenged the effectiveness of conventional detection techniques, motivating the adoption of Graph Neural Networks (GNNs) for their ability to model the structural and semantic information embedded in control flow graphs. While GNNs offer high detection performance, their lack of transparency limits their applicability in security-critical domains. To address this, we present an explainable malware detection framework, which contains a dual explainer. This dual explainer integrates a GNN explainer with a neural subgraph matching approach and the VF2 algorithm. The proposed method identifies and verifies discriminative subgraphs during training, which are later used to explain new predictions through efficient matching. To enhance the generalization of the neural subgraph matcher, we train it using curriculum learning, gradually increasing subgraph complexity to improve matching quality. Experimental evaluations on benchmark datasets demonstrate that the proposed framework retains high classification accuracy while significantly improving interpretability. By unifying explainable graph learning techniques with subgraph matching, the proposed framework enables analysts to gain actionable insights, fostering greater trust in GNN-based malware detectors.
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) | 0.000 | 0.001 |
| Bibliometrics | 0.002 | 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 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
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".