Decomposing and Recomposing Event Structure
Decomposing and Recomposing Event Structure
复制标题
DOI：
10.1162/tacl_a_00445
复制
发表时间：
2021-03
影响因子：
10.9
通讯作者：
William Gantt;Lelia Glass;Aaron Steven White
中科院分区：
文献类型：
--
作者：
William Gantt;Lelia Glass;Aaron Steven White
我们提出了一种事件结构分类，根据句子和文档级通用分解语义（UDS）图注释的推理属性凭经验得出。我们使用由这些图构建的文档级生成模型，将这种分类与语义角色、实体和事件-事件关系分类联合起来。为了支持这一归纳，我们通过一系列推理属性来增强 UDS1.0 数据集中发现的现有注释，该数据集涵盖了整个英语 Web 树库，捕获了事件的时间和方面结构的细粒度方面。生成的数据集（可在 decomp.io 上获取）是迄今为止事件结构和（部分）事件共指的最大注释。
We present an event structure classification empirically derived from inferential properties annotated on sentence- and document-level Universal Decompositional Semantics (UDS) graphs. We induce this classification jointly with semantic role, entity, and event-event relation classifications using a document-level generative model structured by these graphs. To support this induction, we augment existing annotations found in the UDS1.0 dataset, which covers the entirety of the English Web Treebank, with an array of inferential properties capturing fine-grained aspects of the temporal and aspectual structure of events. The resulting dataset (available at decomp.io) is the largest annotation of event structure and (partial) event coreference to date.