Recent research about health trends discussed on Twitter is attracting the attention of public health experts, who believe the social media website could provide insight into public perceptions of health-related issues, NPR’s “Shots” reports.
The study authors are slated to present their findings next week at a conference sponsored by the Association for the Advancement of Artificial Intelligence.
Background
Two computer scientists at Johns Hopkins University’s Center for Language and Speech Processing — Mark Dredze, a professor at the university, and Michael Paul, a graduate student — obtained access to about two billion Twitter updates, or tweets, posted between May 2009 and October 2010 (Barclay, “Shots,” NPR, 7/13).
Dredze and Paul developed a filtering software to pull out and categorize health-related tweets. From their original sample, the two researchers filtered out 1.5 million health-related tweets and sorted them into ailment-specific categories (Armstrong Moore, CNET News, 7/11).
Findings
By analyzing the tweets, the researchers identified Twitter discussions about allergies, insomnia, obesity and numerous other conditions. They also found some correlations between influenza rates and flu discussions on Twitter, as well as allergy-related tweets.
In addition, researchers identified tweets that revealed public misconceptions about health issues. For example, they found tweets suggesting that antibiotics could be used as a flu treatment, even though viruses such as influenza are not susceptible to antibiotics.
Health Experts Show Interest
Paul said that the study has received attention among public health experts who are “interested [in] using Twitter to detect misinformation.” He said such experts view Twitter as a “social media tool to mine people’s perceptions of health.”
Dredze said that analyzing tweets could provide information to help guide public health policies. He said, “It could help officials decide what strategies are effective and what are not” (“Shots,” NPR, 7/13).
Study Limitations
Dredze and Paul’s Twitter study has several limitations, according to CNET News. For instance, some people might lie or stretch the truth in their tweets, and some people might refrain from tweeting more personal health issues.
In addition, the study reflects trends only among people who use Twitter, meaning that certain demographic groups — such as seniors — might not be well-represented (CNET News, 7/11).
Source: iHealthBeat
Comments are closed.
Copyright 2015 - Pulse Practice Solutions | 615.425.2719