Structure-Aware Corpus Construction and User-Perception-Aligned Metrics for Large-Language-Model Code Completion
This work improves evaluation and data processing for code completion in software development, though it appears incremental rather than paradigm-shifting.
The paper addresses the gap between existing evaluation metrics and user perception in code completion tasks by proposing two new metrics (LCP and ROUGE-LCP) and introduces a data processing method (SPSR-Graph) to improve structural semantic modeling for repository-level code completion, demonstrating their effectiveness through experiments.
Code completion technology based on large language model has significantly improved the development efficiency of programmers. However, in practical applications, there remains a gap between current commonly used code completion evaluation metrics and users' actual perception. To address this issue, we propose two evaluation metrics for code completion tasks--LCP and ROUGE-LCP, from the perspective of probabilistic modeling. Furthermore, to tackle the lack of effective structural semantic modeling and cross-module dependency information in LLMs for repository-level code completion scenarios, we propose a data processing method based on a Structure-Preserving and Semantically-Reordered Code Graph (SPSR-Graph). Through theoretical analysis and experimental validation, we demonstrate the superiority of the proposed evaluation metrics in terms of user perception consistency, as well as the effectiveness of the data processing method in enhancing model performance.