Analysing vulnerability reproducibility for Firefox browser
Bibliographic record
Abstract
Fixing some security failures are difficult because they cannot be easily reproduced. To address Hardly Reproducible Vulnerabilities (HRVs), security experts spend a significant amount of time, effort, and budget. Sometimes they do not succeed in the reproduction step and ignore some security failures. The exploitation of a vulnerability due to its irreproducibility may cause severe consequences. An efficient solution is to explore the behaviour of both hardly and easily reproducible security issues at the code level. We use linear regression techniques to build models based on the classical software complexity metrics and a set of attributes related to the environment of the system. The results show that the considered metrics and the vulnerability types do not have significant linear correlations with each other. Also, predicting the HRV-prone parts of large systems is a great help for security experts to focus their effort on the top-ranked vulnerable files. After identifying the suitable indicators based on linear regression, different machine learning techniques such as Random Forest, Logistic Regression, C4.5 Decision Tree, and Naive Bayes are employed to build HRV prediction models. The Random Forest technique achieves the precision of 82% and recall of 84% to classify vulnerable files into HRV-prone or non HRV-prone files. We believe that the results encourage the use of software metrics for vulnerability prediction in some projects.
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 | 0.003 | 0.023 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.006 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".