GeoCoV19

GeoCoV19 是一个大规模的 Twitter 数据集，包含超过 5.24 亿条多语言推文。该数据集包含大约 37.8 万条带有地理标记的推文和 540 万条带有地点信息的推文。注释包括来自用户位置字段和推文内容的地名，并将它们解析为地理位置，例如国家、州或城市级别。在这种情况下，使用用户位置字段对 2.97 亿条推文进行了地理定位注释，使用推文内容对 4.52 亿条推文进行了注释。

GeoCoV19 is a large-scale Twitter dataset comprising over 524 million multilingual Tweets. It contains approximately 378,000 geotagged Tweets and 5.4 million Tweets with associated location information. The dataset's annotations extract place names from both user location fields and Tweet content, which are then resolved into geographic locations at the country, state, or city level. For this dataset, 297 million Tweets were geolocated and annotated using user location fields, while 452 million Tweets were annotated using their content.

搜集汇总

背景与挑战

背景概述

GeoCoV19是一个大规模多语言COVID-19推特数据集，包含超过5.24亿条推文，特别关注地理信息标注，其中37.8万条带有地理标记，540万条包含地点信息。数据集提供了基于用户位置和推文内容的地理位置注释，覆盖国家、州或城市级别。