Naked Data
Issue #276 || Drinking Heavily edition|| 2021-11-29
Subscribe | Archive | About

Map bug reveals notorious Xinjiang camps

China’s internment camps in the Xinjiang province now houses an estimated one million Uighurs and other minorities, arrested for crimes like growing a beard, praying, or downloading WhatsApp. While the rest of the world worries and seethes, China seems immune to eternal criticism. Activists are doing what they can from outside the borders, and one of the most powerful actions they can take is to collect evidence of the mounting human rights abuses.

A Buzzfeed investigation sponsored by the Open Technology Fund, the Pulitzer Center, and the Eyebeam Center for the Future of Journalism, recently made a breakthrough in identifying the detention sites. They discovered a bug in China’s Baidu maps, that showed an anomaly when looking at sensitive sites at a specific zoom level, effectively highlighting the very thing they were meant to hide. “… at a certain point, plain light gray tiles would appear over the camp location. They disappeared as you zoomed in further, while the satellite imagery was replaced by the standard gray reference tiles, which showed features such as building outlines and roads.”

The investigators discovered no less than five million such tiles, since they also masked any area the Chinese deemed sensitive, from power plants to military bases. Through some filtering (the detention centres tend to be near built-up areas) and we assume a lot of hard work, they identified 428 locations. “Of these locations, we believe 315 are in use as part of the current internment program — 268 new camp or prison complexes, plus 47 pretrial administrative detention centers that have not been expanded over the past four years.”

Worryingly, the newly built detention centres are sporting thicker walls as opposed to barbed wire when the programme (or should we say pogrom) started, indicating that China is planning to keep these jails (or should we say concentration camps) open indefinitely. (JNY)

Read about the investigation here.

# Don't Miss

You can judge a burb by its colour

The New York Times data viz team The Upshot takes a different look at predicting which way US voters will swing — the colour of their environment. Green areas (denoting farmland and rural areas) voted Trump, while grey areas (generally urban) voted (unsuccessfully) for Clinton. I’m not sure whether this insight is useful, but it sure is pretty. (JNY)

The numbers that make up the Pacific sex trade

We can almost – almost – forgive the introduction of an interactive wordle into this piece by The Kontinentalist on sex-trafficking in Asia Pacific. It’s an uncomfortable dig into numbers that reveal seven out of ten victims are from the region, using numbers to highlight the systemic weaknesses that make the crime so prevalent. Even the wordle has hidden depths of darkness, as it’s used to navigate a database of media reports on the issue. (AO)

Don’t be a jerk

Tracking the career development of more than 671 students over a decade and a half of their working life is quite the longitudinal study. Nevertheless, Cameron Anderson (a professor at the University of California, Berkley) undertook the mammoth task in order to answer the age old question: do you have to be an asshole to get ahead in business? And the somewhat reassuring answer he came up with is no, you are just as likely to be successful, creative and rise the ladder if you are nice as if you are nasty. (AO)

# Happenings

The state of visualisation

You love dataviz right? And you want to see more and better work in the medium around the world? Then go and help the ever-growing Data Visualization Society map out the state of the industry by taking part in its annual census of datavizzers, and demand some very impressive treatments of the census results. (AO)

# Nerds

Digital museums lean heavily on data viz

Self-confessed museum junky Maxene Graze over at Nightingale takes us through the various stages of museums getting their collections online, from the less-than-satisfactory search to immersive storytelling. The most successful implement the best of data visualisation, and some experiments take data viz much further. (JNY)
Read it here.

Get Tensorflow to write your Tensorflow code

It’s inevitable that machine learning will eventually put data scientists out of work, shortly before The Singularity, after which we’ll all be out of a job, and probably a planet. In the meantime, let’s embrace our soon-to-be overlord and get it to write its own code. With TF-Coder, you give the input and expected output, and it goes and writes the code for you. Yeah, this isn’t going to end well… (JNY)

# Finally

The drinking habits of the South African

Two-thirds of South Africans won’t have missed booze during the long dry months of prohibition. So says fact-checking site Africa Check which backed up health minister Zweli Mkhezi’s claim that only 31% of locals drink this week. The minority of those who do partake in a dram or two more than make up for the abstinence of their peers: they rank a staggering sixth in the world for annual consumption per capita. Check the stats yourself here. (AO)

Who the hell is Jason Norwood-Young?

Journalist, developer, community builder, newsletter creator and international man of mystery, Jason was one of the first South Africans to really grasp the importance of data in the newsroom and has remained one step ahead of the trends in the field all the way. Even Naked Data was conceived before email newsletters were cool again. But what does that tell you about the measure of the man? Nothing, that's what. He hides the superman CV behind a truly mild-mannered and overly modest persona and is best described as "one of the nicest guys in the business". When he's angry, it is righteously so, and his anger always wears velvet mittens. The true signs of his genius include the ability to create multilingual puns on demand (witness the alternative Naked Data strap "Putting the heita in to data") and the fact that he offered me a job. (AO)

Who the hell is Adam Oxford?

Adam combines the best features of his Britishness — polite, sincere, witty, and pasty-white — with the best of South Africanness — enthusiastic, involved, socially conscious, underpaid and overly committed. He wears an incredible amount of hats , including the driving force behind Hacks/Hackers Jo’burg; an "innovation agent" with access-to-justice accelerator HiiL; an editor; a technology journalist; a data journalist; and a budding machine learning and AI expert. His latest gig with Naked Data is a natural progression for a gent that brings a sense of excited purpose to all of his myriad projects. (JNY)

Subscribe at || Copyright © 2020 Naked Data || This work is licensed under a Creative Commons Attribution 4.0 International License
facebook twitter