Online Tool Analyzes Tweets To Determine States’ Caloric Balances

| July 31, 2015 |

University of Vermont researchers used an online, interactive tool to study the number of calories U.S. residents consume, as well as how they expend those calories, the Washington Post‘s “Speaking of Science” reports.

Study Details

The researchers used the tool, called the Lexicocalorimeter, to analyze more than 50 million tweets from Twitter users across the country. The tool searches through the tweets to find words related to food and exercise, such as “doughnut” or “treadmill.” The tool then runs the topic-pertinent tweets through an algorithm that ranks the words by:

  • How often they are mentioned; and
  • How they affect individuals’ caloric intakes and expenditures.

To do this, the algorithm uses a ratio of calories consumed versus calories burned. The ratio allows the algorithm to determine a state’s caloric balance.

Tool Offers Window Into Public Health

According to the analysis, Mississippi residents expended the fewest calories, while Colorado residents expended the most.

Chris Danforth, an assistant professor at UVM and an applied mathematician, said, “In many of the states where obesity rates are the highest, the calories being consumed [are] a lot higher than the calories being burned.”

Danforth noted the study has limitations, such as a restricted sample size. Because of these limitations, the tool could not be used to replace current public health monitoring systems, but it could complement more traditional health indicators. He noted, “We certainly don’t know how long [people are] running or how many hot dogs they’re eating, but from a higher level looking down on Earth, you can see what’s going on with people’s health.”

According to “Speaking of Science,” refining algorithms like the one used by the Lexicocalorimeter is a key factor in bolstering large-scale analyses of using social media to monitor public health.

Mark Dredze, a Johns Hopkins researcher who was not involved in the study, said, “Twitter is really useful for learning what people are talking about and what people are doing.” He added, “Exploring that is the first stage. The second stage is developing better algorithms for the types of questions being asked in public health and determining who in public health will benefit from this information” (Bajak, “Speaking of Science,” Washington Post, 7/29).

Article source

Tags: , , , ,

Category: Medical Technology

About the Author ()

Leave a Reply