arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2610.22150cs.AI
收藏

RAG-RST:作为RST解析上下文篇章模式的检索示例

RAG-RST: Retrieved Exemplars as In-Context Discourse Schema for RST Parsing

Illinois Institute of Technology(伊利诺伊理工学院);Touro University(图罗大学)

Mohammadreza Sediqin, Shlomo Engelson Argamon

AI总结:

本研究提出RAG-RST检索增强框架,无需训练即可将核心性F1提升至0.846、关系F1提升至0.788,动态配置表现最优,证明正式语言模式可作为示例提供给基础模型。

摘要:

在修辞结构理论(Rhetorical Structure Theory,RST)中,篇章树的每条边都带有核心性标签(标记哪个语段包含核心内容)和来自功能类别封闭集合的关系标签。这两种标签都不是表层属性:相同的两个子句可根据修辞意图被标注为详述(Elaboration)或背景(Background),二者的边界由标注惯例而非词汇线索确定。因此预测这些标签需要正式的篇章模式,仅接触富含篇章的预训练文本无法提供该模式:仅给定目标教育单元(EDU)对,强大型语言模型(LLM)的核心性F1值为0.794,关系F1值为0.738。我们提出RAG-RST,一种检索增强框架,它将该模式作为上下文示例提供,而非通过参数更新实现,无需训练即可将核心性F1值提升至0.846,关系F1值提升至0.788。我们对比了两种配置:预计算策略(在解析前为所有EDU对标注)和动态策略(在每个归约步骤进行检索,因此到达生成器的示例取决于解析器已确定的归约操作)。两种配置在语段、核心性、关系及完整树指标上均优于现有RST解析器,其中动态变体始终表现最强。检索到的模式呈现两个特性:其效用受示例数量限制,超过50个示例后效用趋于饱和并下降;且效用取决于结构位置。这两点均表明,正式的语言模式可作为示例提供给基础模型,而非训练至其参数中。

英文摘要:

In Rhetorical Structure Theory (RST), each edge of a discourse tree carries a nuclearity label, marking which span holds the central content, and a relation drawn from a closed inventory of functional categories. Neither label is a surface property: the same two clauses can be Elaboration or Background depending on rhetorical intent, and the boundary between them is fixed by annotation convention rather than by lexical cues. Predicting these labels therefore requires a formal discourse schema, and exposure to discourse-rich pretraining text does not supply it: given only a target EDU pair, a strong LLM recovers nuclearity at 0.794 F1 and relation at 0.738 F1. We propose RAG-RST, a retrieval-augmented framework that supplies this schema as in-context demonstration rather than through parameter updates, raising nuclearity to 0.846 F1 and relation to 0.788 F1 with no training. Two configurations are compared: a precomputed strategy that labels all EDU pairs before parsing, and a dynamic strategy that retrieves at each reduce step, so that the exemplars reaching the generator depend on the reductions the parser has already committed to. Both outperform existing RST parsers across span, nuclearity, relation, and full tree metrics, with the dynamic variant consistently strongest. Two properties of the retrieved schema emerge: its utility is bounded in number, saturating and then declining beyond fifty exemplars, and it depends on structural position. Both indicate that a formal linguistic schema can be supplied to a foundation model as a demonstration rather than trained into its parameters.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑