Scientists have recently invented an instrument to divulge deeper into the billions of posts made on Twitter since 2008. The new tool is capable of providing an unprecedented, minute-by-minute view of popularity. The research was carried out by a team at the University of Vermont. The team calls the instrument the Storywrangler.
The team of eight scientists who invented Storywrangler collect about ten percent of all the tweets made every day around the globe. For every day, they break down these tweets into single bits, as well as pairs and triplets, generating frequencies using nearly a trillion words, hashtags, handles, symbols, and emojis. Storywrangler is the first visualization tool to look at one, two, and three-word phrases across 150 different languages.
The online tool, powered by UVM’s supercomputer, gives a powerful lens for viewing and analyzing the rise and fall of stories, words and ideas each day among people worldwide.
In one significant test of the dataset on the Storywrangler, researchers showed that the instrument could be used to potentially predict financial and political turmoil. They observed the percent change in the use of the words like rebellion in various areas of the world. The researchers found that the rise and fall of these terms were highly associated with a change in the well-established index of geopolitical risk for those same places.
Social media brings billions of voices that are constantly commenting and sharing, complaining and attacking, and thereby in all cases, recording about world wars, weird cats, political movements, deadly diseases, favorite soccer stars, and whatnot.
This tool can pave the way for new approaches in journalism, aid in looking for impressive ways to look at natural language processing as well as the development of computational history.
The UVM team uses Twitter to show how chatter on distributed social media can act as a global sensor system providing information of what happened, how people reacted, and what may come next. Other social media platforms can similarly provide insights into fields ranging from racism to employment, emerging health threats, to new memes.
Researchers present a sample from the Storywrangler’s online viewer in the new Science Advances study, with three globally highlighted events, namely, the death of Iranian general Qasem Soleimani; the Black Lives Matter protests after the murder of George Floyd by Minneapolis police, and the beginning of the COVID-19 pandemic.
According to the researchers, the Storywrangler dataset recorded a sudden spike of tweets and retweets with the term “Soleimani” on January 3, 2020, after the United States assassinated the general. It also takes note of the substantial rise of the term coronavirus and the virus emojis during the spring of 2020 as the disease spread as well as increased use of the hashtag “#BlackLivesMatter” on and after George Floyd was murdered. UVM’s Chris Danforth believes that researchers didn’t know to look for that yesterday, but it will now show up in the data and become part of the story using Storywrangler.