用结构化图增强大模型事件问答能力，无需微调
TAG-EQA: Text-And-Graph for Event Question Answering via Structured Prompting Strategies
- 将因果事件图转为自然语言注入提示，辅助推理
- 在TORQUESTRA上平均提升5%准确率，零样本最高增12%
- 适合需要时序/因果推理的问答场景，可灵活适配不同模型
大语言模型在通用语言任务中表现优异，但在需要因果或时间推理的事件类问题上常表现不佳。本文提出TAG-EQA（Text-And-Graph for Event Question Answering），通过将结构化关系转化为自然语言陈述，将因果事件图注入大模型输入。该框架包含九种提示配置，结合三种策略（零样本、少样本、思维链）与三种输入模态（仅文本、仅图、文本+图），系统分析结构化知识在何种情况下有助于推理。在TORQUESTRA基准测试中，相比纯文本基线，TAG-EQA平均提升5%准确率，零样本设置下最高提升12%，图增强型思维链提示时可达18%。尽管性能随模型和配置变化，研究显示因果图可在不微调的前提下有效提升事件推理能力，为提示式问答提供一种灵活的结构编码方式。
原文摘要 · Abstract (English)
Large language models (LLMs) excel at general language tasks but often struggle with event-based questions-especially those requiring causal or temporal reasoning. We introduce TAG-EQA (Text-And-Graph for Event Question Answering), a prompting framework that injects causal event graphs into LLM inputs by converting structured relations into natural-language statements. TAG-EQA spans nine prompting configurations, combining three strategies (zero-shot, few-shot, chain-of-thought) with three input modalities (text-only, graph-only, text+graph), enabling a systematic analysis of when and how structured knowledge aids inference. On the TORQUESTRA benchmark, TAG-EQA improves accuracy by 5% on average over text-only baselines, with gains up to 12% in zero-shot settings and 18% when graph-augmented CoT prompting is effective. While performance varies by model and configuration, our findings show that causal graphs can enhance event reasoning in LLMs without fine-tuning, offering a flexible way to encode structure in prompt-based QA.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。