XGRL-TCP: An Explainable Graph-Based Reinforcement Learning Framework for Test Case Prioritization in CI
Bibliographic record
Abstract
Continuous integration demands fast and trustworthy fault discovery from large, frequently changing test suites.Many test case prioritization (TCP) methods underperform in this setting because they ignore code-test structure, overlook the semantics of changes, and provide little transparency into ranking decisions.XGRL TCP addresses these gaps with a graph attention network over the test-code dependency graph, CodeBERT embeddings for commit diffs and test text, and an attentionaugmented actor-critic reinforcement learner.On the Defects4J, it achieves an Average Percentage of Faults Detected (APFD) 89.3%, a Fault Detection Rate (FDR) of 85.4%, and a Time to First Fault (TTFF) of 6.3 tests.Key contributions include: (i) a unified structural plus semantic state for TCP, (ii) online adaptation across CI cycles, and (iii) built-in explanatory signals from graph and policy attentions.Compared with contemporary methods TCP-TB and TCP-CIC, XGRL-TCP consistently increases APFD/FDR and reduces TTFF across projects and early execution budgets.Explanations highlight influential code regions and neighboring tests that drive each selection, improving auditability and trust during CI.The approach introduces a modest computational cost: 5.6% overhead, which corresponds to about 33.6 s for a 10-minute suite, leaving typical CI time budgets largely intact.
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 | 0.010 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.004 | 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".