Emerging Results on Automated Support for Searching and Selecting Evidence for Systematic Literature Review Updates
Bibliographic record
Abstract
Context: The constant growth of primary evidence and Systematic Literature Reviews (SLRs) publications in the Software Engineering (SE) field leads to the need for SLR Updates. However, searching and selecting evidence for SLR updates demands significant effort from SE researchers. Objective: We present emerging results on an automated approach to support searching and selecting studies for SLR updates in SE. Method: We developed an automated tool prototype to perform the snowballing search technique and to support the selection of relevant studies for SLR updates using Machine Learning (ML) algorithms. We evaluated our automation proposition through a small-scale evaluation with a reliable dataset from an SLR replication and its update. Results: Effectively automating snowballing-based search strategies showed feasibility with minor losses, specifically related to papers without Digital Object Identifier (DOI). The ML algorithm giving the highest performance to select studies for SLR updates was Linear Support Vector Machine with approximately 74% recall and 15% precision. The use of such algorithms with conservative thresholds to minimize the risk of missing papers can already significantly reduce evidence selection efforts. Conclusion: The preliminary results of our evaluation point in promising directions, indicating the potential of automating snowballing search efforts and of reducing the number of papers to be manually analyzed by about 2.5 times when selecting evidence for updating SLRs in SE.
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 | 0.201 | 0.696 |
| Meta-epidemiology (narrow) | 0.002 | 0.002 |
| Meta-epidemiology (broad) | 0.003 | 0.005 |
| Bibliometrics | 0.023 | 0.015 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.014 | 0.011 |
| Open science | 0.005 | 0.007 |
| Research integrity | 0.004 | 0.003 |
| Insufficient payload (model declined to judge) | 0.013 | 0.006 |
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".