MMED

MMED数据集包含25,165篇来自数百个新闻媒体网站（如Yahoo News、Google News、CNN News）的文本新闻文章和76,516张在Flickr社交媒体上分享的图片，这些数据根据412个真实世界事件进行了标注。该数据集旨在探索组织来自不同数据域的专业人士和业余爱好者贡献的异构数据的问题，以及将从一个数据域获得的事件知识转移到异构数据域的问题，从而总结不同贡献者的数据。

The MMED dataset comprises 25,165 textual news articles from hundreds of news media outlets including Yahoo News, Google News, and CNN News, and 76,516 images shared on the Flickr social media platform. All these data are annotated against 412 real-world events. This dataset is designed to investigate two core research topics: the organization of heterogeneous data contributed by both professionals and amateurs across diverse data domains, and the transfer of event knowledge acquired from one data domain to heterogeneous data domains, with the ultimate goal of summarizing the data contributed by different contributors.