MigrationsKB: A Knowledge Base of Migration related annotated Tweets

<strong>MigrationsKB(MGKB)</strong> is a public Knowledge Base of anonymized <strong>Migration</strong> related <strong>annotated</strong> tweets. The MGKB currently contains over <strong>200 thousand</strong> tweets, spanning over 9 years (January 2013 to July 2021), filtered with 11 European countries of <em>the United Kingdom, Germany, Spain, Poland, France, Sweden, Austria, Hungary, Switzerland, Netherlands and Italy</em>. <strong>Metadata</strong> information about the tweets, such as Geo information (<strong>place name</strong>, <strong>coordinates</strong>, <strong>country code</strong>). <strong>MGKB</strong> contains <strong>entities</strong>, <strong>sentiments</strong>, <strong>hate speeches</strong>, <strong>topics</strong>, <strong>hashtags</strong>, <em>encrypted user mentions</em> in RDF format. The schema of <strong>MGKB</strong> is an extension of TweetsKB for migrations related information. Moreover, to associate and represent the potential economic and social factors driving the migration flows such as <strong>eurostat</strong>, <strong>statista</strong>, etc. FIBO ontology was used. The extracted <strong>economic indicators</strong>, such as GDP Growth Rate, are connected with each Tweet in RDF using geographical and temporal dimensions. The user IDs and the tweet texts are encrypted for privacy purposes, while the tweet IDs are preserved. For this version, the <strong>MGKB</strong> is delivered as a whole and separately by year. The extracted entities and topic words are also published. Online SPARQL endpoint https://mgkb.fiz-karlsruhe.de/sparql/ More information please refer to the website https://migrationskb.github.io/MGKB/. Please contact Yiyi Chen ([e-mail]) for pretrained models (sentiment analysis/hate speech detection/ETM) if necessary.MigrationsKB（MGKB）是一款公开的匿名化迁移相关标注推文知识库。当前MGKB共收录超20万条推文，时间跨度为9年（2013年1月至2021年7月），数据筛选自11个欧洲国家，分别为英国、德国、西班牙、波兰、法国、瑞典、奥地利、匈牙利、瑞士、荷兰及意大利。该数据集包含推文元数据（Metadata），涵盖地理信息，具体包括地点名称（place name）、坐标（coordinates）与国家代码（country code）。MGKB以RDF（Resource Description Framework）格式存储实体（entities）、情感标签（sentiments）、仇恨言论（hate speeches）、主题（topics）、话题标签（hashtags）以及加密用户提及内容。MGKB的模式（schema）是针对迁移相关信息对TweetsKB的扩展。此外，为关联并表征驱动移民流动的潜在经济与社会因素（如欧盟统计局（eurostat）、Statista平台等），研究团队采用了FIBO本体（FIBO ontology）。提取得到的经济指标（economic indicators，如GDP增长率）通过地理与时间维度，以RDF格式与每条推文建立关联。出于隐私保护需求，用户ID与推文文本均已完成加密，仅保留推文ID。本版本的MGKB支持完整包交付与分年度交付两种形式，提取得到的实体与主题词亦同步公开。其在线SPARQL查询端点为https://mgkb.fiz-karlsruhe.de/sparql/，更多信息可访问官方网站https://migrationskb.github.io/MGKB/。若需获取预训练模型（包括情感分析、仇恨言论检测、ETM模型），可联系陈依依（邮箱：[e-mail]）。