Dataset for emergency distortion information identification

The data file consists of four parts. The first part is the publicly available dataset used in this article. The dataset introduction and access links are shown below; The second part is a list of emergency keywords compiled based on the national standard (GB/T 35561-2017); The third part is the experimental data set MEE_data and Test_data after preprocessing, keyword screening and semantic de duplication, as well as the typical case data mig_data of four types of emergencies collected from Zhihu, Xiaohongshu and Tiktok platforms in the subsequent migration test; The fourth part is the appendix of this article, which displays the prompt templates and answer error types of the big language model. The introduction and access links to the public dataset are as follows: 1. Binedata data set, derived from the competition data of Zhiyuan&Computing Institute Internet False News Detection, can be accessed by:“ https://www.biendata.xyz/competition/falsenews/data ”； 2. Rumors dataset, established by authors such as Liu Zhiyuan, accessed through:“ https://opencsg.com/datasets/MagicAI/Chinese_Rumor_Dataset ”； 3. The access method for the CHECK dataset is:“ https://github.com/cyang03/CHECKED ”； 4. CHEF dataset, access method:“ https://github.com/THU-BPM/CHEF ”； 5. MRML dataset, accessed through:“ https://github.com/plw-study/MRML ”； 6. Weibo (Z) dataset, established by Xueyao Zhang et al., accessed through:“ https://github.com/RMSnow/WWW2021 ”； 7. Weibo (M) dataset, established by authors Jing Ma et al., accessed through:“ https://github.com/RMSnow/WWW2021 ”； 8. Jiaozhen data set, an article project from the NLP bus on WeChat official account, can be accessed by:“ https://github.com/lqhou/NLP_Bus_Pro ”； 9. The WBV1 dataset and WBV2 dataset were both established by authors Yang Z and others, and the access method is:“ https://www.kaggle.com/datasets/yangzhou32/socialnet-weibo-version-2 ”.