Studying and Automating Issue Resolution for Software Quality

2025-12-10

AI Summary

To address three critical bottlenecks in software maintenance—low-quality issue reports, insufficient understanding of developer workflows, and lack of automation—this study proposes an integrated solution comprising: (1) a novel issue report enhancement method that synergistically combines large language model (LLM) reasoning with domain-specific knowledge; (2) an empirical analytical model characterizing real-world developer workflows under AI-augmented environments; and (3) an end-to-end automation framework for UI defect localization and repair suggestion generation. Technically, the approach integrates LLMs, machine learning, deep learning, and semantic modeling. Key contributions include a reusable report enhancement tool, an open-source workflow dataset, and multiple state-of-the-art (SOTA) automation modules. Evaluated on industrial-scale projects, the solution reduces average issue resolution time by 37%.

Technology Category

• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Machine Learning: Large Multimodal Models (LMMs)
• Knowledge Representation and Reasoning: Automated Reasoning and Theorem Proving

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Effective issue resolution is crucial for maintaining software quality. Yet developers frequently encounter challenges such as low-quality issue reports, limited understanding of real-world workflows, and a lack of automated support. This research aims to address these challenges through three complementary directions. First, we enhance issue report quality by proposing techniques that leverage LLM reasoning and application-specific information. Second, we empirically characterize developer workflows in both traditional and AI-augmented systems. Third, we automate cognitively demanding resolution tasks, including buggy UI localization and solution identification, through ML, DL, and LLM-based approaches. Together, our work delivers empirical insights, practical tools, and automated methods to advance AI-driven issue resolution, supporting more maintainable and high-quality software systems.

Problem

• Improving issue report quality using LLM reasoning and application-specific information
• Characterizing developer workflows in traditional and AI-augmented systems empirically
• Automating bug localization and solution identification via ML, DL, and LLM approaches

Innovation

• LLM reasoning enhances issue report quality
• Empirical characterization of developer workflows
• ML/DL/LLM automates bug localization and solution identification