A Review of Emotion Detection in Social Media During Covid-19

Last two years since 2019 have proven to be very dreadful where a lot of lives where lost due to covid-19 pandemic. The Covid-19 pandemic proved introduced a lot of anxiety and uncertainty among the population. People shared a lot of things on the social media which proved to be helpful as well as panicky. Due to good thoughts on social media the general population found hope in the devastating pandemic, but with fake social media posts it also created panic among general population due to which a lot of lives where lost due to panic. So social media posts have to be monitored so that correct sentiment of the Covid-19 pandemic can be understood. So, a project has been designed with the help of latest technologies such as text mining, natural language processing and machine learning together. In our project we are going to use twitter as our social media platform as lakhs of tweets are posted daily and can give correct insight of the pandemic and other posts related to it. We are developing the application using a combination of java and python languages. First in our application we will fetch real time tweets using Twitter API. Then the incoming tweets will be cleaned. The cleaned tweets will then be more refined using stop words removal method. Then the stop words removed tweets will then be passed through lemmatization and only words with grammar will be kept. Thus, the final tweets after lemmatization can be used for machine learning. We will design a training dataset used for machine learning using various mediums available on the internet. Then finally we will apply machine learning algorithms SVM and Naïve Bayes and get correct classification of the tweets in two classes positive and negative. Thus, by using the analysis we will be able to understand the correct sentiment behind the tweets and understand emotions behind the Covid-19 pandemic.