Personal Events in Dialogue Corpus

PEDC 是 14 集 This American Life 播客记录的语料库，已针对事件进行了注释。语料库包含这些片段（在表 1 中列出）的摘录，它们是对话。这个语料中标注的粒度就是token；每个标记要么被注释为事件，要么被注释为非事件。有关更多信息，请下载语料库，并查看注释指南以了解有关我们如何定义事件的更多细节，以及有关如何编码注释的 README。此外，有关语料库的更多信息，其用途是从对话纸中自动提取个人事件。

PEDC is a corpus of podcast transcripts from 14 episodes of *This American Life*, annotated for events. The corpus contains dialogue excerpts from the segments listed in Table 1. The annotation granularity of this corpus is at the token level: each token is annotated as either an event or a non-event. For more details, please download the corpus and refer to the annotation guideline for our definition of events and the README document explaining the annotation coding process. Additionally, for more information about the corpus, its intended use is to automatically extract personal events from conversational texts.

搜集汇总

背景与挑战

背景概述

Personal Events in Dialogue Corpus 是一个用于事件抽取的语料库，基于14集This American Life播客的对话内容进行token级别的事件标注。该数据集旨在支持从对话中自动提取个人事件的研究。