paperbot · PL 论文追踪

全部论文 待读 收藏

P2 A unified framework to learn program semantics with graph neural networks

基于图神经网络的程序语义学习统一框架
ASE Proceedings of the 35th IEEE/ACM International Conference on Automated Software Engineering2020匹配度 3/10 · 论文属于用GNN学习程序语义的AI4SE工作,虽涉及类型推断、漏洞检测等程序分析相关任务,但与用户核心关注的“LLM/Agent与静态分析交叉、浮点程序分析、贝叶斯程序分析”交集有限。其方法非LLM/Agent驱动,也未聚焦ML系统或Agentic程序中的静态分析。引用 2 inbox
Shangqing Liu
crossref: 2026-09-12

AI 速览

一句话:提出统一框架,构造两类图以捕获丰富代码语义,服务多种软件工程应用。

程序语义学习是克隆检测、代码摘要等 AI for SE 应用中的关键问题。基于图神经网络(GNN)的程序表示已在漏洞识别、类型推断等任务上取得领先性能,但现有工作缺乏统一框架,且多数方法忽略节点间的全局关系,限制了模型学习丰富语义。本文提出一个统一框架,构造两类图来捕获代码语义,以服务多种软件工程应用。摘要未报告具体实验数据与结果,主要贡献在于统一建图视角与框架设计。

问题
现有基于 GNN 的程序语义学习缺乏统一框架,且多数方法忽略节点间全局关系,难以学习丰富语义。
方法
提出统一框架,构造两类图来捕获代码语义,面向多种软件工程应用。
评估
摘要未详述具体评估设置与结果;提及可应用于克隆检测、代码摘要、漏洞识别、类型推断等任务。
原文摘要(Abstract)

Program semantics learning is a vital problem in various AI for SE applications e.g., clone detection, code summarization. Learning to represent programs with Graph Neural Networks (GNNs) has achieved state-of-the-art performance in many applications e.g., vulnerability identification, type inference. However, currently, there is a lack of a unified framework with GNNs for distinct applications. Furthermore, most existing GNN-based approaches ignore global relations with nodes, limiting the model to learn rich semantics. In this paper, we propose a unified framework to construct two types of graphs to capture rich code semantics for various SE applications.

链接与引用

DOI 原文 · PDF(开放获取) ·

BibTeX
@inproceedings{paperbot5912,
  title = {A unified framework to learn program semantics with graph neural networks},
  author = {Shangqing Liu},
  booktitle = {ASE},
  year = {2020},
  doi = {10.1145/3324884.3418924}
}