Automatic classication of large changes into maintenance categories
Bibliographic record
Abstract
Large software systems undergo significant evolution during their lifespan, yet often individual changes are not well documented. In this work, we seek to automatically classify large changes into various categories of maintenance tasks - corrective, adaptive, perfective, feature addition, and non-functional improvement - using machine learning techniques. In a previous paper, we found that many commits could be classified easily and reliably based solely on the manual analysis of the commit metadata and commit messages (i.e., without reference to the source code). Our extension is the automation of classification by training machine learners on features extracted from the commit metadata, such as the word distribution of a commit message, commit author, and modules modified. We validated the results of the learners via 10-fold cross validation, which achieved accuracies consistently above 50%, indicating good to fair results. We found that the identity of the author of a commit provided much information about the maintenance class of a commit, almost as much as the words of the commit message. This implies that for most large commits, the Source Control System (SCS) commit messages plus the commit author identity is enough information to accurately and automatically categorize the nature of the maintenance task.
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 | 0.015 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.007 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
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".