Semantic Source Code Segmentation using Small and Large Language Models

TL;DR

This paper presents automated methods for segmenting R and Python source code into coherent parts using small and large language models, demonstrating that smaller models fine-tuned on domain-specific data outperform larger pre-trained models.

Contribution

It introduces two novel domain-specific approaches for code segmentation, along with a new annotated dataset, and compares their effectiveness across different language models.

Findings

Context-based line-by-line analysis outperforms range-based segmentation.

Smaller models like CodeBERT and CodeT5+ are more effective than larger LLMs.

Fine-tuning small models on limited annotated data yields superior results.

Abstract

Source code segmentation, dividing code into functionally coherent segments, is crucial for knowledge retrieval and maintenance in software development. While enabling efficient navigation and comprehension of large codebases, manual and syntactic analysis approaches have become impractical as repositories grow, especially for low-resource languages like R and their research domains (e.g., social sciences, psychology).This paper introduces an automated, domain-specific approach for research R code segmentation using Large and Small Language Models (LLMs/SLMs). It presents two novel approaches and a human-annotated dataset, StatCodeSeg. We explore two distinct approaches: line-by-line analysis with context and range-based segment determination. We experiment with LLMs and fine-tuned SLMs. To support the generalizability of our approaches, we also include experiments on Python code from the computer science domain.Our results show that context-based line-by-line analysis is superior over range-based segmentation.Using smaller language models like CodeBERT and an encoder-only version of CodeT5+ are better than their LLM counterparts. Most notably, these two best-performing models did not see R code during pre-training versus the LLMs but were only fine-tuned on 4,130 lines of manually annotated code.