TWEET-FID

TWEET-FID是首个公开可用的社交媒体数据集，专门设计用于支持多种食源性疾病事件检测任务。该数据集由食品科学专家精心策划，覆盖了食源性疾病事件的多层次信息。数据集内容包括从Twitter收集的英文帖子，基于扩展的关键词列表筛选，涵盖了食品、症状、位置和与食源性疾病相关的关键词。创建过程中，通过众包和专家标注相结合的方式进行标注，确保了标注的质量和效率。TWEET-FID的应用领域主要集中在食源性疾病爆发检测，旨在通过自动化提取尽可能多的食源性疾病事件特定信息，加速机器学习模型的开发，以减少疾病爆发的风险和影响。

TWEET-FID is the first publicly available social media dataset specifically designed to support multiple foodborne disease event detection tasks. This dataset was meticulously curated by food science experts, covering multi-level information related to foodborne disease incidents. It consists of English posts collected from Twitter and filtered via an extended keyword list that encompasses terms related to food, symptoms, geographic locations, and foodborne diseases. During its development, the annotation work was conducted through a combination of crowdsourcing and expert labeling, ensuring both annotation quality and efficiency. The primary application domain of TWEET-FID focuses on foodborne disease outbreak detection, aiming to automatically extract as much task-specific information about foodborne disease events as possible to accelerate the development of machine learning models, thereby reducing the risks and impacts of disease outbreaks.