Sentiment Analysis Dashboard of COVID-19 tweets

This Project is part of IBM Hack Challenge 2020. It is a visualization dashboard Webapp imitating descriptive statistics and summaries of COVID-19 tweets data.

All the whereabouts and working of the project and app can be found in this video and this presentation.

Webapp is deployed here.

Built With

• Dash
• Plotly

Data

Data for this app is taken from here due to inavailability of twitter API

Features

• Minimal and Cool UI
• Qualitative analysis with plots
• Lightweight (fast loading speed)
• Supports user interactions

Running App locally

Linux/MacOS

• To be updated...

Windows

• Make sure you have Python3 on your path.
• Clone the repository

  git clone https://github.com/SmartPracticeschool/SBSPS-Challenge-4564-Sentiment-Analysis-Dashboard-of-COVID-19-tweets.git

1. (optional) create and activate new virtualenv or conda env:

pip install virtualenv
virtualenv venv
source venv/bin/activate

or, with conda:

conda create --yes -n dash_docs
source activate dash_docs

2. pip install -r requirements.txt
3. python app.py

Limitations

• This app is not yet responsive for all screen sizes and that's why while loading, it throws an alert for the same.
• As there was no access to twitter's API, the data is taken from web and sentiment of tweets used was precomputed using textblob.
• This app would be updated manually per week, as there is no automation script yet for the tasks.

Screenshots