An Arabic Egyptian Dialect COVID-19 Twitter Dataset (ArECTD)

Abstract

Citizens are increasingly expressing their ideas and feelings on social media platforms such as Twitter. During the coronavirus crisis, numerous emotions are exposed, including sadness, anger, fear, sympathy, surprise, etc. The Arabic Egyptian Dialect COVID-19 Twitter Dataset (ArECTD), comprised of 78K tweets, was collected in the period from the 1st of January 2020 till the 30th of May 2021 focusing on the Egyptian dialect. It was annotated using a combination of manual and a semi-supervised self-learning technique. The tweets of ArECTD were categorized into 10 emotions (sarcasm, sadness, anger, fear, sympathy, joy, hope, surprise, love, and none). Emotion analysis of this dataset could help decision makers understand and respond to the public reactions during the pandemic.