Some interesting data-crunching research projects
A new report considers the demographics of selfies in five cities around the world: Bangkok, Berlin, Moscow, New York and Sao Paolo.
20-30,000 images from each city were analysed both automatically and then using ‘human judgement’ to consider the artistic merits of the images. Mechanical Turk workers were asked to guess the age and gender of the people. After that automatic face analysis created algorithmic estimates of facial positions and emotional expressions. These tags were checked by humans.
Key findings of the report:
- Gender differences – more women than men take selfies (but Moscow has by far the highest female/male gender gap with four and a half times as many selfies being taken by women)
- Selfies are a young person’s pastime – the median age of a selfie taker around the world is 23.7 years.
- Women tilt their heads more than men!
- More people smile in Bangkok; fewer smile in Moscow
Foul language on Twitter
Researchers at Wright State University in the US analysed a random collection of 51 million tweets from 14 million users to measure the ubiquity and context of swearing on Twitter. They identified – amongst other things – different contexts for swearing and which day of the week sees the most swearing. The full PDF of the article can be downloaded by a link on this page – but beware. The article uses swear words throughout and is probably not suitable for work!
Pizza and value for money
Analysis of almost 75,000 pizza prices in the US sets out to answer once and for all the tricky question of pizza size versus value for money.
Using data to predict the Oscars
BuzzFeed’s Data Scientist seems to have successfully predicted many of the winners and losers in this piece (published before the ceremony). He correctly predicted a Best Picture win for 12 Years a Slave, a win for Matthew McConaughey and best original screenplay for Her.)
Pizzas and selfies at the Oscars
Of course pizzas and selfies were combined at this year’s Oscar ceremony when the host and several Hollywood A-listers posed for what was to become one of the most retweeted images ever – the ‘selfie to beat all selfies’. They then ordered in pizza.