STAR-ML: A Rapid Screening Tool for Assessing Reporting of Machine Learning in Research
Bibliographic record
Abstract
Literature review provides researchers with an overview of the field and when presented as a systematic assessment, it summarizes state-of-the-art information and identifies knowledge gaps. While there are many tools for assessing quality and risk-of-bias within studies, there is currently no generalized tool for evaluating the transparency, reproducibility, and correctness of machine learning (ML) reporting in the literature. This study proposes a new tool (Screening Tool for Assessing Reporting of Machine Learning; STAR-ML) that can be used to screen articles for a systematic or scoping review focusing on the reporting of the ML algorithm. This paper describes the development of the tool to assess the quality of ML research reporting and how it can be applied to improve the literature review methodology. The tool was tested and updated using three independent raters on 15 studies. The inter-rater reliability and the time used to review an article were evaluated. The current version of STAR-ML has a very high inter-rater reliability of 0.923, and the average time to screen an article was 4.73 minutes. This new tool will allow for filtering ML-related papers that can be included in a systematic or scoping review by ensuring transparent, reproducible, and correct screening of research for inclusion in the review article.
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 | 0.582 | 0.840 |
| Meta-epidemiology (narrow) | 0.005 | 0.005 |
| Meta-epidemiology (broad) | 0.009 | 0.016 |
| Bibliometrics | 0.085 | 0.048 |
| Science and technology studies | 0.004 | 0.004 |
| Scholarly communication | 0.014 | 0.019 |
| Open science | 0.006 | 0.023 |
| Research integrity | 0.005 | 0.005 |
| Insufficient payload (model declined to judge) | 0.022 | 0.008 |
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; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
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".