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 2307.01972cs.CL
收藏

基于增量提示与验证的开放域层次事件模式归纳

Open-Domain Hierarchical Event Schema Induction by Incremental Prompting and Verification

University of Illinois at Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校);University of Pennsylvania(宾夕法尼亚大学)

Sha Li, Ruining Zhao, Manling Li, Heng Ji, Chris Callison-Burch, Jiawei Han

AI总结:

本文提出将事件模式视为可从大型语言模型推导的常识知识,并设计增量提示与验证方法分三阶段构建复杂事件图,在时序和层次关系F1上分别提升7.2%和31.0%,且生成模式的可读性和事件覆盖率优于现有模型。

摘要:

事件模式是关于事件典型进展过程的一类世界知识。近期的事件模式归纳方法使用信息抽取系统从文档中构建大量事件图实例,然后从这些实例中学习归纳出模式。与之不同,我们提出将事件模式视为一种可从大型语言模型(LLMs)中推导出的常识知识。这一新范式极大地简化了模式归纳过程,并使我们能够以直接的方式同时处理事件之间的层次关系和时序关系。由于事件模式具有复杂的图结构,我们设计了一种增量提示与验证方法,将复杂事件图的构建分解为三个阶段:事件骨架构建、事件扩展以及事件-事件关系验证。与直接使用LLMs生成线性化图相比,我们的方法能够生成更大且更复杂的模式,在时序关系上F1提升7.2%,在层次关系上F1提升31.0%。此外,与先前最先进的封闭域模式归纳模型相比,人类评估者在将模式转化为连贯故事时能够覆盖约多10%的事件,并在可读性方面(5分制)给我们的模式打出高1.3分的评分。

英文摘要:

Event schemas are a form of world knowledge about the typical progression of events. Recent methods for event schema induction use information extraction systems to construct a large number of event graph instances from documents, and then learn to generalize the schema from such instances. In contrast, we propose to treat event schemas as a form of commonsense knowledge that can be derived from large language models (LLMs). This new paradigm greatly simplifies the schema induction process and allows us to handle both hierarchical relations and temporal relations between events in a straightforward way. Since event schemas have complex graph structures, we design an incremental prompting and verification method to break down the construction of a complex event graph into three stages: event skeleton construction, event expansion, and event-event relation verification. Compared to directly using LLMs to generate a linearized graph, our method can generate large and complex schemas with 7.2% F1 improvement in temporal relations and 31.0% F1 improvement in hierarchical relations. In addition, compared to the previous state-of-the-art closed-domain schema induction model, human assessors were able to cover $\sim$10% more events when translating the schemas into coherent stories and rated our schemas 1.3 points higher (on a 5-point scale) in terms of readability.

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。

↑