Measuring global food sentiment one tweet at a time
Why
While food (literally) keeps us alive it also gives us massive sensory enjoyment. It forms a cornerstone of cultural activity. From new unions and family reunions to remembering our national heritage we use food to celebrate important occasions. Conversely a bad meal can taint an experience. Lack of food has devastating effects. Food choices provide clues to people’s personal philosophies and possibly their status in the community.
Globally, food has has become the rising star of public debate. The food movement has generated unprecedented focus on the subject. We have never been more aware of what we’re eating and the effect it has on our bodies and the planet. Developed countries see conscious citizens turning to the slow food and organic movements. Yet, obesity levels are set to reach record highs of 2.3 billion in 2015 (WHO). Google Zeitgeist shows the popularity of food and drink searches on its engine has exponentially increased over the last decade. Arguably, citizen interest, consumer focus and the food industry itself are currently at an enormous tipping point.
Aim
Do Happy Meals make Americans happy? Do the British find comfort in cake? Are citizens of wealthy countries enjoying food more than those from poorer nations? Does national history manifest itself in present-day food consumption? And what is bun cha* anyway?
Food Mood aims to gain a better understanding of global food consumption patterns and its impact on the daily emotional well-being of people against the backdrop of countries’ GDPs and obesity levels.
Process
Food Mood is an interactive information visualization and sentiment analysis tool using Twitter as a primary data source. It is a collaborative project between Affect Lab and the brilliant people at AI Applied. FoodMood is a work in progress: find the beta site at www.foodmood.in.
Using natural language processing techniques, all English language tweets are collected and analyzed for sentiment, location and foodstuff. We capture, in real time, the foods that people are tweeting about and how they feel about them.
We then add a layer of public access data from the World Health Organization (country obesity levels) and country GDP (CIA Factbook) we can develop an even more informed picture about food and emotion.
Throw in the geo-located data of the tweets and it gets juicy!
Results
The constantly evolving findings indicate:
- The most enjoyed foods per country
- Foodstuff and sentiment comparisons between countries
Previous findings include:
- Meat enjoys a high sentiment rating and is vastly consumed the world over
- Fast-food companies dominate the food consumption landscapes of most countries and yet not all of them enjoy equal sentiment ratings across countries.
- As the adage goes, “money doesn’t make you happy” and the same could be said for food, as the sentiment analysis of the poorest countries of the world indicates in this project.
Why data visualization?
Data visualization is a powerful technique for taking a vast quantity of data and making it easy to understand in one glance. This is the challenge we faced with Food Mood. Collecting all the data was the easy part. Showing the connections, patterns and relationships that exist between them is the part that data visualization helped resolve.
* Grilled Pork Noodle Soup (Vietnamese: Bún chả) is a Vietnamese dish, which is thought to be originated from Hanoi, the capital of Vietnam. Bun cha is served with a plate of white rice noodle (bún) and herbs in a steamy broth.
