Studying Twitter

Posted 13th September 2017 by Mawsley
A number of extreme American libertarian and right-wing news websites have fumed over the news that a scientist has been given almost $200,000 of public money to study vaping on Twitter. While it may be an odd use of cash, studying mentions of ecigs on Twitter is not new to the lead author or the academic research community.

People are questioning “the Feds” wasting $199,665 on getting some people to sit and look at tweets on the social media platform. The proposal for the study claimed that the ban on sales to young people meant “surveillance of evolving themes and factors contributing to message popularity for e-cigarette chatter on social media platforms is an important activity”.

The plan is to study Twitter and Reddit and monitor what type of phrases are retweeted, upvoted, or replied to. The reasoning is: “We expect these results will help health agencies, the FDA, and researchers gain insights into observed viral nature of certain messages and designing effective strategies to maximize diffusion of their messages.”

Ramakanth Kavuluru is the man behind the successful proposal, a computer scientist at the University of Kentucky. Kavuluru specialises in word and data mining along with information extraction.

The thing that is notable is the crossover from his previous research work. In 2016, Kavuluru was singing his own praises, in a paper titled Toward automated e-cigarette surveillance: “we believe our work demonstrates the strong potential of informatics approaches, specifically machine learning, for automated e-cig surveillance on Twitter.”

Then, earlier this year, he co-produced an almost identical paper titled On quantifying diffusion of health information on Twitter, but this was more concerned with perfecting a methodology: “We need a way to identify missing nodes that might make the retweet graphs more connected.”

Previously, at the tail end of 2015, researchers from Drexel University in California and Glantz’ University of Southern California combined to look at tweets about Blu electronic cigarettes. They concluded, after spending a lot of money, that Twitter could be used to spread marketing messages.

This fed into a later study by the Schroeder Institute for Tobacco Research and Policy Studies. Their work followed a similar style and discovered: “Twitter has become the ‘wild-west’ of marketing and promotional strategies for advertisement agencies.” At least the graphs were prettier.

Researchers at the University of San Diego actually used Twitter in a different fashion. They felt anonymously listening in on conversations could provide and answer to the question of why vapers vape. They concluded: “Our data suggest the reasons people vape are shifting away from cessation and toward social image. The data also show how the ENDS [vape] market is responsive to a changing policy landscape. For instance, smoking indoors was less frequently cited in 2015 as indoor smoking restrictions became more common.”