Skip to content

ETL of large corpus of COVID related tweets, used to train sentiment ML model and ridge regression correlational model.

License

Notifications You must be signed in to change notification settings

logan-life/covid_cases_big_twitter_data

Repository files navigation

Introduction

The World Well-Being Project is an inter-disciplinary effort among computer scientists, psychologists, and statisticians. Based out of the University of Pennsylvania’s Positive Psychology Center and Stony Brook University’s Human Language Analysis Lab, the WWBP develops scientific and programmatic methods for measuring physical and psychological health using social media data.

We were inspired by the great work that the WWBP does in the health outcomes space. For example, the researchers at the WWBP has successfully demonstrated that machine learning models trained on twitter data are able to predict heart disease mortality rates and excessive alcohol use.

Following in that theme, we were interested to investigate if there is a correlation between twitter data and COVID case rates. In the end, we did find a strong correlation between TFIDF scores for twitter ngrams + COVID case rates, as well as sentiment scores and COVID case rates.

Instructions

The Twitter Terms of Service prevent us from providing the full tweet json objects publicly. To run this notebook, you need that data. Please contact us if you wish to discuss using our data.

The easiest way to use this code is to run the notebook in Google Collab and upload the data from this repository to your Google Drive account. The notebook will then walk through the process step-by-step.

About

ETL of large corpus of COVID related tweets, used to train sentiment ML model and ridge regression correlational model.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published