HALO : a multi-feature two-pass analysis to identify framework API evolution
Bibliographic record
Abstract
ABSTRACT: Software frameworks and libraries are indispensable to today's software systems. Because of the fast development of open-source software in recent years, frameworks and libraries have became much versatile as any open-source system or part thereof can be used as a framework (or a library). Developer can reuse frameworks in innovative ways that are not expected by the providers of frameworks. Many frameworks are not well documented and very few owners provide specific documents to describe the changes between different releases of their frameworks. When they evolve, it is often time-consuming for developers to keep their dependent code up-to-date. Approaches have been proposed to lessen the impact of framework evolution on developers by identifying API evolution or change rules between two releases of a framework. However, the precision and recall of the change rules generated by these approaches depend on the features that they use, such as call-dependency relations or text similarity. If these features do not provide enough information, the approaches can miss correct change rules and compromise the precision and recall. For example, if a method in the old release of a framework is not called by other methods, we cannot find its change rule using call-dependency relations alone. Considering more features can overcome this limitation. Yet, because many features may also give contradictory information, integrating them is not straightforward. We thus introduce Halo, a novel hybrid approach that uses multiple features, including call dependency relations, method documentations, inheritance relations, and text similarity. Halo implements a two-pass analysis inspired by pattern classification problem. We implement Halo in Java and compare it with four state-of-the-art approaches. The comparison shows that, on average, the recall and the precision of Halo are 43% and 5% higher than that of other approaches.
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 | 0.007 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.010 | 0.004 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.003 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.002 |
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".