COVID19_sentinentanalysissocialmedia

• The global pandemic that has raged the world for the past year is here to stay. In this repo, we try to predict the various sentiments associated with a particular tweet after training on readily available datasets, the links to which have been attached below.
• The multi-label benchmark dataset can be retrieved from the first reference down below.

• To access the BERT model: https://www.dropbox.com/s/jr4mdva8no5w7v1/bertmodel.pth?dl=0

References

• SenWave: Monitoring the Global Sentiments under the COVID-19 Pandemic, Yang, Qiang and Alamro, Hind and Albaradei, Somayah and Salhi, Adil and Lv, Xiaoting and Ma, Changsheng and Alshehri, Manal and Jaber, Inji and Tifratene, Faroug and Wang, Wei and others https://arxiv.org/pdf/2006.10842.pdf (Note: If you want to use the labled tweets, please mail to qiang.yang[AT]kaust[dot]edu[dot]sa to get the pwd for the zip filefolder.)
• Semeval-2018 Task 1: Affect in Tweets. Saif M. Mohammad, Felipe Bravo-Marquez, Mohammad Salameh, and Svetlana Kiritchenko. In Proceedings of the International Workshop on Semantic Evaluation (SemEval-2018), New Orleans, LA, USA, June 2018.
• Tweets originating from India during Covid-19 Lockdowns 1, 2, 3, 4 - https://ieee-dataport.org/open-access/tweets-originating-india-during-covid-19-lockdowns-1-2-3-4