An exploration of selfies

Posted to Data Art  |  Tags: , ,  |  Nathan Yau

Selfiecity, from Lev Manovich, Moritz Stefaner, and a small group of analysts and researchers, is a detailed visual exploration of 3,200 selfies from five major cities around the world. The project is both a broad look at demographics and trends, as well as a chance to look closer at the individual observations.

There are several components to the project, but Imageplots (which you might recognize from a couple years ago) and the exploratory section, aptly named Selfiexploratory, will be of most interest.

Selfie imageplots

The two parts let you filter through cities (Bangkok, Berlin, Moscow, New York, and Sao Paulo), age, gender, pose, mood, and a number of other factors, and this information is presented in a grid layout that self-updates as you browse.

So you can get a rough sense of how facets relate. There seems to be a higher proportion of female selfies and average age seems to skew towards younger as you’d expect. The average age of females in this selfie sample seems to be younger than that of males.

However, before you jump to too many conclusions about how countries vary or differences between the sexes, etc, consider the classification process, which was a combination of manual labor via Mechanical Turk and face recognition software. Age, for example, can be though to estimate from pictures alone since you have outside factors like makeup, angles, and poses. Do these things account for the two- to three-year average difference between the sexes? Maybe. So consider the data. But that should go without saying.

That said, Selfiecity is a fun one I spent a good amount of time browsing. It’s a weird, tiny peek into 3,200 people’s lives, with a dose of quant and art. And don’t miss the theoretical component in essay format, a reflection of social media, communities, and the self.

Favorites

Pizza Place Geography

Most of the major pizza chains are within a 5-mile radius of where I live, so I have my pick, …

Causes of Death

There are many ways to die. Cancer. Infection. Mental. External. This is how different groups of people died over the past 10 years, visualized by age.

The Most Unisex Names in US History

Moving on from the most trendy names in US history, let’s look at the most unisex ones. Some names have …

Divorce and Occupation

Some jobs tend towards higher divorce rates. Some towards lower. Salary also probably plays a role.