Covid-19-Sentiment-Analysis

Covid-19 Tweets Sentiment Analysis for IBM Hack Challenge 2020

Check out our hosted -> app
Check out the -> video demonstration of the app

About IBM HACK Challenge

CODE FOR A BRIGHTER TOMORROW

IBM Hack Challenge 2020 is all about coding for a cause. It’s about finding solutions to problems plaguing society today. It is an opportunity to showcase your coding talent, learn new technologies and build an efficiently working solution.

About our problem statement

Sentiment Analysis of COVID-19 Tweets Visualization Dashboard

The sentiment analysis of Indians after the extension of lockdown announcements to be analyzed with the relevant #tags on twitter and build a predictive analytics model to understand the behavior of people if the lockdown is further extended. Also develop a dashboard with visualization of people reaction to the govt announcements on lockdown extension.

What is Sentiment Analysis?

Sentiment Analysis is a technique widely used in Data analysis. Twitter Sentiment Analysis, therefore means, using data analysis techniques to analyze the sentiment of the tweets, thus classifying tweets into the categories of positive, negative and neutral.

In this project, we have used the VADER Sentiment Analysis tool for classifying tweets into positive, negative or neutral polarity.

Our Approach

• Collect Tweets 📚
  • Used GetOldTweets3 for collecting historical Twitter data 💹
• Preprocess / Clean Tweet data 🧹
  • Remove RTs, URLs, Non-English Tweets
• Perform Sentiment Analysis on Tweets' text 😌
  • Classify Tweets into
    • Positive polarity ❤️
    • Negative polarity 😢
    • Neutral polarity 😐
• Organize data for different plots 😅
  • Day wise tweet sentiment scores for day-wise analysis of rate of change of people sentiment
  • Month wise tweet sentiment scores for month-wise analysis of rate of change of people sentiment
  • Phase wise tweet sentiment scores for sentiment analysis during various phases (Lockdown 1.0 - 4.0 & Unlock 1.0,2.0)
  • City wise tweet sentiment scores for analyzing sentiment across various cities
  • Frequency of hashtags for creating wordcloud
  • Overall sentiment score data for plotting overall sentiment plots
• Make beautiful visualizations (graphs and plots!) 😍
  • Doughnut plot
    
  • Bar charts
    
  • Geographical plots
    
  • Wordclouds
    
  • Hashtag plots
    
  • And More!
    
• Deploy the visualization dashboard online for the world to see! 😏
  • Optimizing entire data for a smooth user-experience on the app
    
    

Gifs of hosted app!

Project report and presentation

They are stored in the docs/ folder of this repo

• Check out the
• or the

Authors

1. Animesh Singh

2. Aaditya Kapoor