fake news detection
MultiClaim-2025-es
- Read more about MultiClaim-2025-es
- Log in or register to post comments
Este conjunto de datos consiste en verificaciones de hechos, publicaciones de redes sociales y emparejamientos entre ambos. Incluye 205.751 verificaciones de hechos en 39 idiomas y 28.092 publicaciones en redes sociales en 27 idiomas. Todas las publicaciones han sido previamente revisadas por verificadores profesionales, que también les asignaron las verificaciones correspondientes. Hay 31.305 pares de verificación-publicación, y cada publicación está asociada con al menos una verificación. De estos pares, 26.774 son monolingües y 4.212 son multilingües.
CheckThat Lab Task 2 on Claim Normalization Monolingual
NLP topic
Dataset
Language
Spanish
English
Arabic
Deuch
French
Hindi
Portuguese
Year
2025
CheckThat! Lab Task 3 on Fact-Checking Numerical Claims
NLP topic
Dataset
Language
Spanish
English
Arabic
Year
2025
mCLAN-2025-es
- Read more about mCLAN-2025-es
- Log in or register to post comments
The dataset contains claims from social media and fact checking websites, with their normalized version.
CheckThat!Numerical-2025-es
- Read more about CheckThat!Numerical-2025-es
- Log in or register to post comments
The dataset contains multigenre content in Arabic, English, and Spanish. The dataset is collected from various fact-checking domains through Google Fact-check Explorer API2, complete with detailed metadata and an evidence corpus sourced from the web.
PAN-AP-2020-ES
- Read more about PAN-AP-2020-ES
- Log in or register to post comments
FakeDeS
- Read more about FakeDeS
- Log in or register to post comments
CT–CWT–22-ES
- Read more about CT–CWT–22-ES
- Log in or register to post comments
CheckThat-ES
- Read more about CheckThat-ES
- Log in or register to post comments
CT–CWT–23-ES
- Read more about CT–CWT–23-ES
- Log in or register to post comments
The dataset focused on three topics: COVID-19, climate change and technology. The Spanish dataset is a combination of CT-CWT-21, CT-CWT-22 and newly collected content. It is composed of tweets collected from Twitter accounts and transcriptions from Spanish politicians, which are manually annotated by professional journalists who are experts in fact-checking. Each tweet was labeled using both the image and the text.
