- This repository contains the COVIDEmotions annotated data.
- Each tweet has at least 2 annotations and at most 5 annotations.
- There are a total of 903 annotations.
- Each row of the training and development set contains a unique tweet ID and is followed by its labeled emotions. The content of tweets in our dataset is anonymized for privacy. The Twitter API can be used to collect the specific content of the tweets using their unique tweet ID.
alexandertekle / covidemotions Goto Github PK
View Code? Open in Web Editor NEWCOVIDEmotions