---
title: "针对GNN知识图谱问答的查询侧攻击：从实体链接到答案生成追踪失败环节"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2608.25922v1"
permalink: "https://feed.astralor.com/items/itm_adfa59686db9f8b74250c538"
published: "2026-08-26T15:34:29.000Z"
score: 74
channel: "ai"
tags: ["ComplexWebQuestions","WebQSP","图神经网络","对抗攻击","知识图谱问答"]
---

# 针对GNN知识图谱问答的查询侧攻击：从实体链接到答案生成追踪失败环节

## AI 点评

> 这篇论文提醒我们，端到端指标在复杂管线中可能会严重误导我们对薄弱环节的判断。对于正在构建可解释、可维护的KGQA系统的人来说，子图构建才是值得下功夫的地方，而不是一味堆模型复杂度。

## 速览

该论文提出一种阶段隔离协议，用于定位GNN知识图谱问答管线在对抗扰动下的失效环节。通过两类保持答案的扰动（组合重构与关系同义词替换），在ComplexWebQuestions和WebQSP上评估后发现：GNN推理阶段在子图完整时性能几乎不变，而子图构建环节对端到端性能崩溃的贡献超过99%。即使74%的检索子图中存在正确答案，子图构建仍导致答案无法被触及，说明答案存在性与答案可达性是本质不同的概念，缓解应聚焦于子图构建而非推理模型。

## 关键点

- 提出阶段隔离协议，区分实体链接、子图检索、GNN推理、答案生成四阶段的失败贡献
- 两类对抗扰动组合重构（CR）与关系同义词替换（RS）在保持答案的前提下针对不同阶段
- 实验发现子图构建是主要崩溃源，贡献超过99%的端到端性能下降，而GNN推理阶段保持近基线准确率
- 74%的失败检索子图实际包含正确答案，揭示答案存在性与可达性之间的关键差异

## 全文

GNN-based Knowledge Graph Question Answering (KGQA) pipelines process queries through four discrete stages: entity linking, subgraph retrieval, GNN reasoning, and answer generation. Standard robustness evaluations conflate stage-level failures into a single end-to-end metric, obscuring both the source of brittleness and the appropriate mitigation target. We ask which stage fails, and why, when the pipeline is subjected to adversarial perturbations on the input question. We introduce a stage-isolation protocol with two answer-preserving adversarial perturbations verified against the knowledge graph: Compositional Restructuring (CR) and Relation Synonym Swap (RS) target distinct stages while leaving entity seeds intact. Evaluated across ComplexWebQuestions and WebQSP, the results run counter to prevailing assumptions: the GNN reasoning stage retains near-baseline accuracy when the subgraph is intact, while subgraph construction accounts for over 99\% of the end-to-end collapse under CR, occurring even when the gold answer is present in 74\% of retrieved subgraphs. This exposes a fundamental distinction between answer presence and answer reachability that end-to-end metrics cannot detect, and places the mitigation target firmly at the subgraph construction stage rather than the reasoning model. Perturbed datasets and evaluation infrastructure are released at https://anonymous.4open.science/r/atkgrag-E85C .

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_adfa59686db9f8b74250c538