Oi, #tubby! You are what you Tweet, boffins find

Boffins from the USA and Australia have constructed a “Lexicocalorimeter” that parses social media for mentions of food, discovering a correlation between people's wellbeing, and the foods and activities they mention on Twitter.

In an ArXiv paper titled Gauging public health through caloric input and output on social media (PDF), the authors explain that a Lexicocalorimeter is an instrument “for measuring the 'caloric content' of social media and other large-scale texts.”

The instrument's been fed “tables of food and activity related phrases”, and those terms have been matched to “sourced estimates of caloric intake and expenditure.” The Lexicocalorimeter then analysed 50 million geotagged tweets to count “the total number of times each food and physical activity phrase in our database was tweeted about.”

With that data in hand, the authors estimated how many calories folks in each American state consume and expend, creating a “caloric balance” that measures whether residents are likely to be burning all the food they ingest.

Lexicocalorimeter data was then compared to “health and well-being demographics for the contiguous United States”.

Their conclusion: States whose residents mention lots of lard online are also full of overweight and sick people.

In other words, you are what you Tweet.

The most-Tweeted foods in the USA. Click here to embiggen.

The paper mentions several similar studies that have reached similar conclusions, and concludes by suggesting analysis of a more comprehensive vocabulary - soft drinks weren't analysed this time around - could assist with all manner of public health analysis challenges. ®