Using Social Media to Track Global Eating Habits

According to researchers, social media can be used to track and understand eating habits from around the world.  These researchers also proposed that the data about eating habits could then be used to “identify cultural boundaries and similarities across populations”. Computer scientists at the University of Birmingham along with members from the Department of Computer Science, Universidade Federal de Minas Gerais in Brazil analyzed geospatial data pulled from 500,000 users of Foursquare to understand temporal differences in eating and drinking habits.

Foursquare is a social media platform where users check in to specific locations and share tips and comments about different venues.  The researchers wanted to understand the ‘temporal’ habits of people relating to food, for example, the distribution of meals throughout the day, where those meals were being eaten, what types of venues are popular, and whether there is any correlation in the eating habits between different cities and countries.

They found that Brazilians eat their biggest meal at lunchtime, those in the United Kingdom save the biggest feast for dinnertime, and people in the United States tend to eat all hours of the day (although dinner in the US is also the biggest meal of the day).

Researchers also found that France and Brazil have a strong correlation in drinking habits while the eating habits of Londoners are similar to those in New York and San Francisco.  These three cities also share similarly diverse and international aspects such as all three have a Chinatown district.

Dr Mirco Musolesi, lead investigator from the University of Birmingham’s School of Computer Science, said: ‘Eating and drinking is an essential part of the culture and characteristics of a country and can be used to understand how cities are composed of communities of different nationalities.

‘The availability of social media datasets allows us to understand the culture and the habits of people at planetary level and at a granularity that is really unprecedented. Traditionally, these studies have been conducted at a much smaller scale, and have involved a very small sample of people. This is because they are based on questionnaires or interviews and, they are extremely expensive and time-consuming.

‘My research lab at Birmingham is investigating the possibility offered by these new data sources, including mechanisms that are able to preserve the privacy of individuals as much as possible. The applications are many: not only for researchers, but also for retail, commerce, industry, government and no-profit organisations. We are really just at the beginning.’

Reference