NYU Data Science newsletter – April 4, 2016

GROUP CURATION: N/A

Anyone upset about wasted government spending should take a look at for-profit scientific publishing, a $25 billion industry that double-dips into taxpayers’ pockets. In 2014 one of the largest firms, the Anglo-Dutch Elsevier, which publishes Cell and the Lancet, made a profit of £762 million (about $1.2 billion) on gross revenue of £2 billion (about $3.1 billion)—a margin of 37%. The problem is that much of the research in these journals is supported by federal money from the National Science Foundation and the National Institutes of Health. Then the journals charge taxpayers for the right to see the research their money paid for.

This has long been the promise of new technology: That it will make your work easier, which will make your life better. The idea is that the arc of technology bends toward social progress. This is practically the mantra of Silicon Valley, so it’s not surprising that Google’s CEO, Sundar Pichai, seems similarly teleological in his views.

“Hundreds of years ago very few people had access to information. And they were essentially in the corridors of power,” Pichai told Mat Honan, a long-time technology reporter and BuzzFeed’s San Francisco bureau chief. “Even a simple thing like the printing press made books accessible to many more people. I’ve always been fascinated by this thing, that every jump in technology involves leveling the playing field.”

During the conference, nearly 900 messages were posted on Twitter which included the hashtag #pyastro16. On this page we present an overview of the social activity by displaying, in chronological order, the 164 “most popular” tweets which were retweeted or favourited more than five times.

There’s an opportunity around location that extends beyond targeting consumers with ads for mayonnaise or corn flakes when they’re walking down an aisle in a grocery store.

“The location data opportunity is still unfolding,” said Tracey Scheppach, EVP of precision video at Publicis Groupe’s Starcom MediaVest. “As agencies start to get smart about things like that, the data will be used for more than just placing ads; it’ll be used to inform business decisions. The role of the advertising agency is starting to change as awareness goes up around using data to drive more than just media decisions.”

Some of the most significant inroads in this area have been around attribution.

… My approach has been not to reduce one to the other, or to create some kind of hybrid; my approach has been parallel, where I attempt to do both well but not necessarily together.

You have already started to answer my next question, which pertains to the role of the digital in your humanities work. Do you think this qualifies as “digital humanities” work? And do you care?

I definitely don’t care. I felt very much involved in the digital world before digital humanities came on the scene. When it happened, digital humanities for me seemed like a reboot of some of the things that had been happening in information science and library schools years earlier. I feel as if I am doing digital work but not really digital humanities. But I have the luxury of not needing to care. In my reading of it, DH started in English, comparative literature, and language departments. My degree is technically in literature, but I work in a media studies program. I basically studied theory in graduate school, not literature, so I was never in the pressure-cooker environment that English and language departments have vis-à-vis jobs and what counts as research. I mean what more can you really say about Shakespeare today? There isn’t a whole lot. But if you start counting words, then maybe there is something new you can say. You see this frequently in very old, extremely erudite, well-established disciplines where there is very little territory left in which to do research.

Our lives have become inundated by recommendation engines. On a daily basis, we receive recommendations on things to do, purchase, and watch from the likes of Amazon, Netflix, and Yelp. At face value, we take recommendations for granted—most recommendations simply make sense: people who like this one action movie will probably like this other action movie too. But there’s a lot going on under the hood that makes them work, sometimes with remarkable insight. … When you think about this like a data scientist, you have your user, and then you have your items, and each of those has a list of associated properties (features). One thing you can do is look at the properties the user likes—based on items they have purchased in the past or items they have reviewed or rated—as well as the properties of each item, and then start matching them that way. This is what’s referred to as content-based filtering.

A community led collection of recipes, build infrastructure and distributions for the conda package manager.