Progressive Code Integration for Abstractive Bug Report Summarization

TL;DR

This paper introduces a progressive code-integration framework for LLM-based bug report summarization, effectively combining code snippets and text to produce more accurate and comprehensive summaries, outperforming traditional methods.

Contribution

It presents a novel incremental approach that integrates code snippets with textual bug reports, overcoming LLM context limits and improving summary quality.

Findings

Outperforms extractive baselines by 7.5%-58.2%

Achieves comparable performance to state-of-the-art abstractive methods

Enhances bug report comprehension by jointly leveraging text and code

Abstract

Bug reports are often unstructured and verbose, making it challenging for developers to efficiently comprehend software issues. Existing summarization approaches typically rely on surface-level textual cues, resulting in incomplete or redundant summaries, and they frequently ignore associated code snippets, which are essential for accurate defect diagnosis. To address these limitations, we propose a progressive code-integration framework for LLM-based abstractive bug report summarization. Our approach incrementally incorporates long code snippets alongside textual content, overcoming standard LLM context window constraints and producing semantically rich summaries. Evaluated on four benchmark datasets using eight LLMs, our pipeline outperforms extractive baselines by 7.5%-58.2% and achieves performance comparable to state-of-the-art abstractive methods, highlighting the benefits of jointly leveraging textual and code information for enhanced bug comprehension.