Sentiment-Analysis-Covid-19

Comparitive Sentiment Analysis of deep learning models for Covid-19 tweets dataset. This project compares performance of 4 different deep learning models on a covid-19 tweets dataset that was created using twitter API. The 4 models are - CNN, LSTM, CNN - LSTM, LSTM - CNN. Out of these 4 it was found that LSTM performed the best on a 12k tweet dataset with 79.9 % accuracy, LSTM - CNN with 51.61 %, CNN - LSTM with 49 % and CNN with 46.6 %. Hashtags used were - Corona, Covid - 19, CoronaVirus, Lockdown, EconomicCrisis, MakeChinaPay, Pandemic, NatureIsHealing, Recession2020, SocialIsolation, StayHomeSaveLives, WeAreTheVirus, StayHome, Quarantine, LockdownDiaries, LiquorShopsOpen, EndTheLockdown, ExtendLockdown.