5803 tweet pairs in the French language human-annotated for automatic stance detection. The tweets were collected during July and August 2018. Four classes broadly adopted by the community are used for the annotation: Support, Deny, Query, and Comment with the addition of the Ignore class. The corpus along with the tools used to build it are made available.
In compliance with Twitter terms of use, we do not share the content of the tweets, but only the tweets IDs.
Please fill in the agreement form and indicate the name of the corpus (Event2018) in your application.
The corpus is available at the INA scientific dataset website.
You can then unzip the file tweet_annots.json.zip
.
You can find more info about the Event2018 dataset here: https://github.com/ina-foss/twembeddings.
Some results of this Tweets Stance corpus analysis and the software tools to build it and to perform these analyses will be available in the coming days.
Web-based application developed under flask allowing for the annotation of tweet stance. Two levels of annotation are offered, the stance of the target tweet toward the tweet it responds to, and toward the root tweet of the thread. The database (not provided - for Twitter privacy policy reason) is in the .sqlite format.