News / Vancouver

SFU student using big data to find world's top chocolate-loving cities

Saif Charaniya analyzed Instagram data to determine which city takes the cake for international cocoa capital, and the answer is surprisingly nutty.

People in �sk�dar, a municipality of Istanbul, Turkey, post the highest number of chocolate-related images to Instagram, according to research carried out by Simon Fraser University student Saif Charaniya.

Flickr: Judy van der Velden

People in �sk�dar, a municipality of Istanbul, Turkey, post the highest number of chocolate-related images to Instagram, according to research carried out by Simon Fraser University student Saif Charaniya.

A Simon Fraser University student analyzing Instagram data to find the world’s top chocolate-loving cities says he has made a nutty discovery.

While some may think Brussels, famed for its Belgian chocolates, or Paris, known for its chic candy shops, would take the cake for having the highest-number of chocolate-related posts with location tags on the popular photo-sharing platform, Saif Charaniya said the data reveals an unexpected answer.

According to his analysis, the title of international cocoa capital by a significant margin goes to Üsküdar, a municipality of Istanbul, Turkey.

Simon Fraser University masters student Saif Charaniya, a self-proclaimed chocolate addict, is using big data from Instagram to find the world’s top chocolate-loving cities.

Contributed

Simon Fraser University masters student Saif Charaniya, a self-proclaimed chocolate addict, is using big data from Instagram to find the world’s top chocolate-loving cities.

“That was very surprising,” Charaniya, a self-proclaimed chocolate addict, told Metro. “I was expecting it to be New York or L.A., just because that’s where most of the Instagram users live.”

For his research, Charaniya collected mounds of data— 40 gigabytes worth— over a period of five weeks. After finding 9.4 million posts that included the hashtag #chocolate, Charaniya then narrowed the data down to images that were also tagged with a location, finding 1.2 million posts.

Not only did Üsküdar score highest for chocolate-related posts with a location tag, the city also came out on top for the highest number of likes per post and highest number of comments per post.

Charaniya also uncovered other surprising trends.

Globally, Valentines Day was the second most popular day of the year for posting about chocolate, with the most popular day being Easter, he said.

In Brazil, the numbers of posts about chocolate on Valentine’s Day was lower than other days of year, he added, while in Russia, chocolate posts went up on International Women’s Day.

In Vancouver, the number of chocolate posts spiked by 319 per cent on Valentine’s Day, the third highest increase globally after Tokyo and Denver, which rose by 410 per cent and 362 per cent, respectively.

Toronto took the ninth spot, with a 228-per-cent jump in the number of chocolate posts on the holiday of love.

Some candy brands, like Snickers, also appear to enjoy more brand loyalty online, with a higher number of comments on photos than any other chocolate brand.

“It’s interesting,” said Charaniya. “These anomalies are what we should focus on in further research to understand what encourages people to talk about chocolate.”

Charaniya said the research underscores the importance of using social media to harness and analyze big data for companies keen on taking a bite out of the $98-billion dollar global chocolate market.

“People love chocolate— it’s a worldwide phenomenon,” he said. “If you look at the way that people talk about chocolate, you can understand how users consume chocolate, what are their favourite brands and what occasions do they talk about it.”

A student in SFU’s big data masters program, Charaniya said his next goal is to use big data to measure public sentiment by detecting and understanding how an audience is reacting to an event, brand or news story.

“Social media is one of the biggest contributors to big data,” he said. “You can find out so much about a person just by looking at their Facebook, Twitter or Instagram.”

See the rest of Charaniya's reseach here.