LinkRank: A Learning-to-Rank Framework for One-to-Many Issue-Commit Traceability

• URL: http://arxiv.org/abs/2607.23610v1
• Date: Sun, 26 Jul 2026 11:33:51 GMT
• Title: LinkRank: A Learning-to-Rank Framework for One-to-Many Issue-Commit Traceability
• Abstract summary: LinkRank is a learning-to-rank framework for recovering one-to-many issue--commit links.<n>LinkRank considers the set of candidate commits for an issue and identifies the commits that are most likely to contribute to its resolution.<n>We evaluate LinkRank under two settings: Known-K, where the true number of linked commits is available, and Unknown-K, where the model must infer when to stop selecting commits.
• Score: 11.169387830830628
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Recovering traceability links between issues and commits is important for software maintenance, debugging, impact analysis, and project understanding. However, most existing approaches assume a one-to-one relationship, where each issue is linked to a single commit. In practice, many issues are resolved through multiple commits, and ignoring this one-to-many nature can lead to incomplete traceability. This paper presents LinkRank, a learning-to-rank framework for recovering one-to-many issue--commit links. Unlike existing methods that mainly judge issue--commit pairs independently, LinkRank considers the set of candidate commits for an issue and identifies the commits that are most likely to contribute to its resolution. To support realistic evaluation, we construct a new dataset from six open-source GitHub repositories. LinkRank follows an iterative pick--remove--renormalize strategy: it selects the highest-ranked commit, removes it from the candidate pool, renormalizes the remaining scores, and repeats the process until the stopping criterion is met. We evaluate LinkRank under two settings: Known-K, where the true number of linked commits is available, and Unknown-K, where the model must infer when to stop selecting commits using ABS and REL stopping rules. Across six projects, LinkRank achieves an average Known-K F1 score of 74.54%, compared with 48.39% for the strongest baseline. In the Unknown-K setting, LinkRank achieves 68.84% F1 with ABS and 67.02% F1 with REL, outperforming the strongest baselines under both automatic stopping rules. Overall, the findings suggest that one-to-many issue--commit traceability is better addressed as an issue-centric ranking and iterative selection problem than as independent pairwise classification.

Related papers

• Disagree to Explore, Agree to Commit: Routing-Guided Test-Time Scaling for Software Agents [60.650808962786364]
  We study whether native MoE router traces can guide steering and selection without an external judge or selection-time test execution.<n>Our analysis shows that routing provides a robust behavioral role signal; token-granular readouts and decision-matched comparison sets turn it into effective control.<n>We evaluate on SWE-bench Verified using open-weight sparse MoE agents across scales and reasoning-effort settings.
  arXiv  Detail & Related papers  (2026-08-23T03:07:03Z)
• Entity Resolution in Practice: Lessons from a Self-Serve Pipeline [0.39146761527401436]
  We built and evaluated a self-serve entity resolution (ER) system on six benchmarks spanning 864 to 5M records.<n>We recommend training several algorithm families per dataset and letting an automatic bake-off pick the winner.<n>One false-positive link can silently merge unrelated entities.
  arXiv  Detail & Related papers  (2026-07-28T21:41:32Z)
• Don't Commit Alone: Joint Token Commitment in Diffusion Large Language Models [0.6916773850242582]
  We propose CoCommit, a marker-gated coordination pass that briefly defers commitment.<n>On LLaDA2.1-mini with LoRA adapters and matched greedy inference, joint commitment improves accuracy on all six benchmarks we evaluate.
  arXiv  Detail & Related papers  (2026-07-05T19:29:13Z)
• CORE-T: COherent REtrieval of Tables for Text-to-SQL [91.76918495375384]
  CORE-T is a scalable, training-free framework that enriches tables with purpose metadata and pre-computes a lightweight table-compatibility cache.<n>Across Bird, Spider, and MMQA, CORE-T improves table-selection F1 by up to 22.7 points while retrieving up to 42% fewer tables.
  arXiv  Detail & Related papers  (2026-01-19T14:51:23Z)
• Programming over Thinking: Efficient and Robust Multi-Constraint Planning [54.77940831026738]
  SCOPE is a framework that disentangles query-specific reasoning from generic code execution.<n>SCOPE achieves state-of-the-art performance while lowering cost and latency.
  arXiv  Detail & Related papers  (2026-01-14T02:58:07Z)
• CARE What Fails: Contrastive Anchored-REflection for Verifiable Multimodal [84.71254539482369]
  Group-relative reinforcement learning with verifiable rewards (RLVR) often wastes the most informative data it already has the failures.<n>We present CARE, a failure-centric post-training framework for multimodal reasoning that turns errors into supervision.<n> CARE improves accuracy and training smoothness while explicitly increasing the share of learning signal that comes from failures.
  arXiv  Detail & Related papers  (2025-12-22T16:34:21Z)
• LinkAnchor: An Autonomous LLM-Based Agent for Issue-to-Commit Link Recovery [1.5399429731150376]
  A study on GitHub shows that only 42.2% of the issues are correctly linked to their commits.<n>We present LinkAnchor, the first autonomous LLM-based agent designed for issue-to-commit link recovery.
  arXiv  Detail & Related papers  (2025-08-17T04:21:44Z)
• LLM-Driven Collaborative Model for Untangling Commits via Explicit and Implicit Dependency Reasoning [15.20947984949809]
  We propose ColaUntangle, a new collaborative consultation framework for commit untangling.<n>ColaUntangle integrates Large Language Model (LLM)-driven agents in a multi-agent architecture.<n>We evaluate ColaUntangle on two widely-used datasets (1,612 C# and 14k Java tangled commits)
  arXiv  Detail & Related papers  (2025-07-22T09:42:13Z)
• DCR: Divide-and-Conquer Reasoning for Multi-choice Question Answering with LLMs [9.561022942046279]
  We propose Divide and Conquer Reasoning (DCR) to enhance the reasoning capability of large language models (LLMs) We first categorize questions into two subsets based on confidence score ($mathcalCS$), which is estimated by statistical frequency of generated answers. In particular, we first categorize questions into two subsets based on confidence score ($mathcalCS$), which is estimated by statistical frequency of generated answers.
  arXiv  Detail & Related papers  (2024-01-10T14:38:46Z)
• Delving into Commit-Issue Correlation to Enhance Commit Message Generation Models [13.605167159285374]
  Commit message generation is a challenging task in automated software engineering. tool is a novel paradigm that can introduce the correlation between commits and issues into the training phase of models. The results show that compared with the original models, the performance of tool-enhanced models is significantly improved.
  arXiv  Detail & Related papers  (2023-07-31T20:35:00Z)
• Automated Recovery of Issue-Commit Links Leveraging Both Textual and Non-textual Data [2.578242050187029]
  Current state-of-the-art approaches for automated commit-issue linking suffer from low precision, leading to unreliable results. We propose Hybrid-Linker to overcome such limitations by exploiting two information channels. We evaluate Hybrid-Linker against competing approaches, namely FRLink and DeepLink on a dataset of 12 projects.
  arXiv  Detail & Related papers  (2021-07-05T09:38:44Z)
• Chained-Tracker: Chaining Paired Attentive Regression Results for End-to-End Joint Multiple-Object Detection and Tracking [102.31092931373232]
  We propose a simple online model named Chained-Tracker (CTracker), which naturally integrates all the three subtasks into an end-to-end solution. The two major novelties: chained structure and paired attentive regression, make CTracker simple, fast and effective.
  arXiv  Detail & Related papers  (2020-07-29T02:38:49Z)
• TACRED Revisited: A Thorough Evaluation of the TACRED Relation Extraction Task [80.38130122127882]
  TACRED is one of the largest, most widely used crowdsourced datasets in Relation Extraction (RE) In this paper, we investigate the questions: Have we reached a performance ceiling or is there still room for improvement? We find that label errors account for 8% absolute F1 test error, and that more than 50% of the examples need to be relabeled.
  arXiv  Detail & Related papers  (2020-04-30T15:07:37Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.