MobIE

MobIE是一个德语数据集，专注于移动领域的命名实体识别、实体链接和关系抽取。该数据集包含3,232个社交媒体文本和交通报告，总计91,000个tokens，并标注了20,484个实体，其中13,104个实体与知识库链接。数据集的创建过程涉及从2015年到2019年收集的德语Twitter消息和RSS feeds，并通过人工和弱监督方法进行标注。MobIE的应用领域主要集中在交通和公共运输问题，旨在通过多任务学习提升信息抽取的准确性。

MobIE is a German-language dataset dedicated to named entity recognition, entity linking, and relation extraction in the mobile domain. It encompasses 3,232 social media texts and traffic reports, totaling 91,000 tokens, with 20,484 annotated entities, of which 13,104 entities are linked to external knowledge bases. The dataset was developed by collecting German Twitter posts and RSS feeds from 2015 to 2019, and annotated using both manual and weak supervision methods. MobIE is primarily focused on traffic and public transportation applications, with the objective of enhancing the accuracy of information extraction through multi-task learning.

搜集汇总

背景与挑战

背景概述

MobIE是一个德语数据集，包含3,232个文档，涵盖Twitter消息和交通报告，标注了细粒度地理实体、标准命名实体以及交通相关的关系和事件。该数据集适用于命名实体识别、实体链接和关系提取任务，支持多格式访问。