A recent report declared Wyoming as the most toxic state on Twitter, but a University of Wyoming professor feels that claim might be dubious.
An artificial intelligence program used as a business writing assistant at Writer.com has determined the tweets posted by Wyoming residents are among the most toxic in the nation. As of Monday morning, 17% of Wyoming’s tweets were considered “toxic.”
Writer defined “toxic” as aggressive, mean, offensive, racist, sexist or hateful commentary.
But a University of Wyoming communications and journalism professor questioned the report’s validity.
“I checked out the website and it uses machine content analysis to code tweets as toxic or not. Without access to their code book or their machine’s coding of toxic words, it’s hard to know the details about what is going on,” UW professor Dr. Kristen Landreville told Cowboy State Daily. “I agreed that some of their coded tweets as toxic were indeed toxic, but others were just disagreeable rather than toxic.”
The states identified as the most toxic by Writer change from day to day. Last week, Wyoming was in the top 10, but on Monday, the state was at a lower ranking.
On Monday, the most toxic state on Twitter was Delaware.
Writer’s AI writing assistant uses natural language processing and deep learning to determine toxicity, including grammar, style, tone, adherence to key terms and core messages, inclusiveness and healthy communication.
Landreville also questioned how many tweets in Wyoming were actually being coded in the research.
‘We could have just a handful of Wyomingites tweets dominating the tweets that this computer/machine is coding,” she said. “They claim to only take a sample of opinion leaders from Wyoming and code them. But there’s no way for us to see who’s tweets are being coded, and since we know Wyoming is small in population, there could be a few people driving this ‘toxic’ label.”
