Bibliographic record
Abstract
This deposit contains four (4) main datasets that were used in the study "Dependabot and Security Pull Requests: Large Empirical Study" (Link). Each dataset is described as follows : Dataset (1) - Dependency Update : This dataset concerns issues related to pull requests (PRs) that were created by both users and bots to manage dependency updates in GitHub projects. The search was based on the keywords "Dependency, Update" in the title, body or comment of a PR created in the time period between 26/05/2017 and 15/06/2021 for the 1st partition, and between 01/01/2023 and 30/09/2023 for the 2nd partition. We obtained a total of 6,573,489 PR-related issues belonging to a total of 927,007 repositories for partition (1); and for partition (2), we obtained a total of 3,342,829 PR-related issues belonging to a total of 816,028 repositories. Dataset (2) - Dependabot Security PRs : The second dataset is related to PRs created by Dependabot to handle security vulnerabilities in project dependencies. In our search, we look for PR-related issues created by "Dependabot-preview" or "Dependabot" and with the label "security", also created during the time period between 26/05/2017 and 30/09/2023. With these parameters, our results consist of 422,388 issues from 47,987 repositories. Dataset (3) - Manual Security PRs : For this dataset, we were interested in PRs created only by users to handle security vulnerabilities. The search consists of finding the keywords "Dependency, Vulnerable" in the title, body or comment of a PR created in the time period between 26/05/2017 and 30/09/2023. We only consider pull requests created by authors with the type "user". The final results include a total of 186,186 issues for 60,758 repositories. Dataset (4) - Bots' Security PRs : This dataset is related to PRs created by several bots to handle security vulnerabilities in project dependencies. In the search query, we look for PR-related issues where the keywords "Dependency", and "Security", and "Vulnerability" are mentioned in the title, body, or comment of the PR. These PRs are created by one of the following bots: "Snyk", "Renovate", "Greenkeeper", or "Depfu", also created during the time period between 26/05/2017 and 30/09/2023. The obtained results for the 4 bots consists of a collection of 628,495 PR-related issues in a total of 105,342 repositories. We also included : Derived Sample : This sample contains the data that was selected and extracted to conduct our manual qualitative analysis, and the manual feature extraction.
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 | 0.016 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.006 | 0.007 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.007 | 0.005 |
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".