HIPE-2022 Shared Task Named Entity Datasets

HIPE-2022 datasets used for the HIPE 2022 shared task on <strong>named entity recognition and classification (NERC) and entity linking (EL) in multilingual historical documents</strong>. HIPE-2022 datasets are based on six primary datasets assembled and prepared for the shared task. Primary datasets are composed of historical newspapers and classic commentaries covering ca. 200 years, feature several languages and different entity tag sets and annotation schemes. They originate from several European cultural heritage projects, from HIPE organizers’ previous research project, and from the previous HIPE-2020 campaign. Some are already published, others are released for the first time for HIPE-2022. The HIPE-2022 shared task assembles and prepares these primary datasets in HIPE-2022 release(s), which correspond to a single package composed of neatly structured and homogeneously formatted files. Primary datasets undergo the following preparation steps: conversion to the HIPE format (with correction of data inconsistencies and metadata consolidation); rearrangement or composition of train and dev splits. Please also refer to: HIPE-2022 shared task website: https://hipe-eval.github.io/HIPE-2022/ HIPE-2022 data repository: https://github.com/hipe-eval/HIPE-2022-data Here is an overview of the primary datasets: <strong>Dataset alias</strong> <strong>Readme</strong> <strong>Document type</strong> <strong>Languages</strong> <strong>Suitable for</strong> <strong>Project</strong> hipe2020 link historical newspapers de, fr, en NERC-Coarse, NERC-Fine, EL CLEF-HIPE-2020 newseye link historical newspapers de, fi, fr, sv NERC-Coarse, NERC-Fine, EL NewsEye sonar link historical newspapers de NERC-Coarse, EL SoNAR letemps link historical newspapers fr NERC-Coarse, NERC-Fine LeTemps topres19th link historical newspapers en NERC-Coarse, EL Living with Machines ajmc link classical commentaries de, fr, en NERC-Coarse, NERC-Fine, EL AjMC The HIPE-2022 team expresses her greatest appreciation to the partnering projects, namely AJMC, <em>impresso, </em>HIPE-2020, <em>Living with Machines</em>, <em>NewsEye</em>, and <em>SoNAR</em>, for contributing their NE-annotated datasets (and hiding a part thereof for the time of the evaluation campaign).

HIPE-2022数据集是用于HIPE 2022共享任务的数据集，该任务聚焦<strong>多语言历史文档中的命名实体识别与分类（Named Entity Recognition and Classification, NERC）及实体链接（Entity Linking, EL）</strong>。HIPE-2022数据集基于为本次共享任务整理筹备的6个核心数据集构建而成。核心数据集涵盖约200年跨度的历史报纸与经典评论文本，涉及多种语言、不同的实体标注集与标注范式，其来源涵盖多个欧洲文化遗产项目、HIPE主办方的既往研究项目，以及往届HIPE-2020评测活动。其中部分数据集已公开发布，其余则为HIPE-2022首次公开释放。

HIPE-2022共享任务将这些核心数据集整合并整理为HIPE-2022发布包，该发布包为单一整合包，包含结构规整、格式统一的各类文件。核心数据集需经过以下预处理步骤：转换为HIPE格式（修正数据不一致问题并统一元数据）；拆分或重组训练集与开发集。

请参阅以下资源：
HIPE-2022共享任务官网：https://hipe-eval.github.io/HIPE-2022/
HIPE-2022数据集仓库：https://github.com/hipe-eval/HIPE-2022-data

以下为核心数据集概览：
<strong>数据集别名</strong> <strong>说明文档</strong> <strong>文档类型</strong> <strong>语言</strong> <strong>适用任务</strong> <strong>所属项目</strong>
hipe2020 链接 历史报纸 德语、法语、英语 粗粒度命名实体识别与分类（NERC-Coarse）、细粒度命名实体识别与分类（NERC-Fine）、实体链接（EL） CLEF-HIPE-2020
newseye 链接 历史报纸 德语、芬兰语、法语、瑞典语 粗粒度命名实体识别与分类（NERC-Coarse）、细粒度命名实体识别与分类（NERC-Fine）、实体链接（EL） NewsEye
hipe2020 链接 历史报纸 德语 粗粒度命名实体识别与分类（NERC-Coarse）、实体链接（EL） SoNAR
letemps 链接 历史报纸 法语 粗粒度命名实体识别与分类（NERC-Coarse）、细粒度命名实体识别与分类（NERC-Fine） LeTemps
topres19th 链接 历史报纸 英语 粗粒度命名实体识别与分类（NERC-Coarse）、实体链接（EL） Living with Machines
ajmc 链接 经典评论文本 德语、法语、英语 粗粒度命名实体识别与分类（NERC-Coarse）、细粒度命名实体识别与分类（NERC-Fine）、实体链接（EL） AjMC

HIPE-2022团队谨向各合作项目致以最诚挚的谢意，包括AJMC、<em>impresso</em>、HIPE-2020、<em>Living with Machines</em>、<em>NewsEye</em>以及<em>SoNAR</em>，感谢它们贡献了经命名实体标注的数据集（并在本次评测活动期间隐藏了部分数据以用于评估）。