Automatically Identifying Solution-Related Content in Issue Report Discussions with Language Models DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; June 03–05, 2018; Woodstock, NY ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Software and
Antu Saha, Mehedi Sun, Oscar Chaparro · 2025
Read Automatically Identifying Solution-Related Content in Issue Report Discussions with Language Models DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; June 03–05, 2018; Woodstock, NY ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Software and by Antu Saha, Mehedi Sun, Oscar Chaparro (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-06501.txt.
Description
Antu Saha, Mehedi Sun, and Oscar Chaparro automate the hunt for solution talk inside software issue discussions. When a bug is debated, the thread mixes proposals, code changes, and evaluations, and finding the solution-bearing parts by hand is slow. Reopened bugs, regressions, and code reviews all depend on locating that content quickly.
Three language model strategies were compared: embeddings, prompting, and fine-tuning, across traditional, pretrained, and large model classifiers in sixty-eight configurations. A dataset from three hundred fifty-six Firefox issues grounded the tests. Embeddings beat word-count features, prompting lagged, and a fine-tuned large model scored best, with ensembles doing better still and transferring to other projects.
Catalog record
| CREATOR | Antu Saha, Mehedi Sun, Oscar Chaparro |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 113 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-06501