SIEVE：面向软件工程研究的参数化 GitHub 语料库构建器
原标题：SIEVE: A Parameterized GitHub Corpus Builder for Software Engineering Research
AI 摘要
论文提出开源工具 SIEVE，一个参数化的 GitHub 语料库构建器，可在函数、方法和类级别从仓库中提取代码记录，支持按编程语言、仓库质量和创建日期筛选。每条记录附带 22 项结构指标、AST 特征、可选的基于 CodeBERT 的 AI 生成概率评分及依赖信息，并兼容 CodeSearchNet 和 OpenClassGen 格式。对 400 条记录的人工验证显示代码指标与导入检测准确率 100%，源码完整性 99.0%，元数据准确率 99.5%。该工具旨在降低软件工程研究中语料构建成本，并支持带时间约束的 LLM 评测以减少数据污染。
正文节选
SIEVE: A Parameterized GitHub Corpus Builder for Software Engineering Research Abstract Software engineering research increasingly relies on large, richly annotated corpora of real-world code. Still, researchers typically assemble these with ad-hoc scripts or static, pre-built datasets that offer no control over corpus composition. This leaves a gap for a reusable tool that goes below the repository level and builds annotated, function- and class-level code corpora under researcher-specified cri