Princeton CITP, Freedom to Tinker blog, Arunesh Mathur
from
In a paper that is set to appear at the 2018 IEEE Workshop on Consumer Protection in May, we conducted a study to better understand how content creators on social media disclose their relationships with advertisers to end-users. Specifically, we examined affiliate marketing disclosures—ones that need to accompany affiliate links—-which content creators placed along with their content, both on YouTube and Pinterest.
How we found affiliate links. To study this empirically, we gathered two large datasets consisting of nearly half a million YouTube videos and two million Pinterest pins. We then examined the description of the YouTube videos and the Pinterest pins to look for affiliate links. This was a challenging problem, since there is no comprehensive public repository of affiliate marketing companies and links.
Researchers trying to understand how neural networks work shouldn’t just focus on interpretable neurons, according to new research from DeepMind researchers.
Kogan graduated from UC Berkeley in 2008 with a degree in psychology. He was part of the Haas Scholars Social Science Fellowship program from 2007 to 2008, for which he carried out research entitled “The Secrets of the Heart: Love Directionality and Construct Integration.”
The project included objectives such as studying the “personality, cultural, and situational factors which contribute to the vast differences individuals have in their definitions of love,” according to the Haas Scholars website.
To collect data on Facebook users for Cambridge Analytica, Kogan used his app, “thisisyourdigitallife,” which offered users personality predictions if they allowed the app access to their personal information, according to Bloomberg. The data was then used to access information about the users’ friends, as well.
Palantir, a secretive company co-founded by billionaire Peter Thiel, worked with Cambridge Analytica, the political analysis firm that harvested data from Facebook users, whistleblower Christopher Wylie told U.K. lawmakers Tuesday.
Wylie claimed that Cambridge Analytica CEO Alexander Nix was introduced to Palantir by Sophie Schmidt, the daughter of former Google CEO Eric Schmidt.
Thiel, who has not been accused of any wrongdoing, was not immediately available for comment when contacted by CNBC.
Wylie, a former employee of Cambridge Analytica, said Nix had “several meetings” with Palantir.
Amid public concerns over Cambridge Analytica’s use of Facebook data and a subsequent movement to encourage users to abandon Facebook, there is a renewed focus on how social media companies collect personal information and make it available to marketers.
Pew Research Center has studied the spread and impact of social media since 2005, when just 5% of American adults used the platforms. The trends tracked by our data tell a complex story that is full of conflicting pressures. On one hand, the rapid growth of the platforms is testimony to their appeal to online Americans. On the other, this widespread use has been accompanied by rising user concerns about privacy and social media firms’ capacity to protect their data.
All this adds up to a mixed picture about how Americans feel about social media. Here are some of the dynamics.
The New York Times, Michael Wines and Emily Baumgaertner
from
At least 12 states signaled Tuesday that they would sue to block the Trump administration from adding a question about citizenship to the 2020 census, arguing that the change would cause fewer Americans to be counted and violate the Constitution.
The New York State attorney general, Eric T. Schneiderman, said he was leading a multistate lawsuit to stop the move, and officials in Connecticut, Delaware, Illinois, Massachusetts, New Jersey, New Mexico, Oregon, Pennsylvania, Rhode Island and Washington said they would join the effort. The State of California filed a separate lawsuit late Monday night.
“The census is supposed to count everyone,” said Attorney General Maura Healey of Massachusetts. “This is a blatant and illegal attempt by the Trump administration to undermine that goal, which will result in an undercount of the population and threaten federal funding for our state and cities.”
One of China’s top tech companies is trying to push the frontiers of artificial intelligence by teaching computers to understand scenes from the 1990’s romance-disaster epic Titanic.
The technology, from China company SenseTime, is supposed to distinguish Titanic’s romantic scenes from disaster scenes. Although most humans would have no problem distinguishing Jack and Rose’s blooming love from the Titantic’s sinking, the feat is highly complex for computers.
In a demonstration at MIT Technology Review‘s EmTech Digital conference in San Francisco on Monday, the technology performed well and was able to classify the scenes correctly. It highlights the advancement of artificial intelligence, but also how far it still has to go before becoming able to understand more complex movie scenes outside of public demonstrations.
Across the military branches, commanders are looking at how artificial intelligence can be used to fly better aerial routes, insert robotic vehicles into formations and process vast amounts of data in the information-dense battlefields of the near future.
The Defense Advanced Research Projects Agency recently started a program that looks to use that same technology and other tools to get inside the enemy’s head, learning their intent in the nebulous “gray zone” of conflict that marks modern warfare.
Gray zone conflicts are those in which state and non-state competition becomes conflict but remains below the level of conventional warfare. Experts have pointed to Russia’s use of hybrid threats in Ukraine and other areas, along with China’s aggression in the South China Sea as examples.
Two years ago in Sun Valley, ID, media moguls were told by Silicon Valley investors that artificial intelligence is coming to businesses “and it’s coming quickly.” Well, it has just arrived for one well-known market research firm — Fizziology, which is now part of Marketcast. Social media research firm Fizziology is implementing artificial intelligence for what is believed to be the first-ever, real-time monitoring using A.I. that will discover sudden changes on social media conversation surrounding a project, brand or ad category.
They are calling it VelocityWatcher and it monitors conversation outside of owned social channels as well to include comedic perspectives, complaints and the like — issues that don’t necessarily appear in automated dashboards.
A group of teenagers has been interrogating me for almost an hour. “Can you prove you are who you say you are?” one of them asks. “Show us the tear stains on your diploma to prove you actually went to U Chicago,” another says. I don’t have the tear stains. But I do have a Twitter account with my name on it, and enough memory of my time at college to offer convincing details that I actually attended the University of Chicago as an undergraduate. Eventually the group provisionally agrees that I am real.
They have good reason to be suspicious: This team of about 40 incoming freshmen, who call themselves the Secret Squids, have spent more than a month investigating something called PS, a secret society that has existed at the University of Chicago for more than a century.
Led by the Reticulites, 11 masked monks clad in red robes, the purpose of PS is to track the Parasite, a mysterious force that appears on campus once every 11 years in the form of a misplaced, miraculously conjured room. Eventually the Squids discovered the unsettling fact that the secret society had planted several fake students among their number: By the time I joined them in August, paranoia had become the norm.
To complete their initiation and become full members of PS, the Squids—as well as everyone else in the 1,800-plus class of 2021—were invited to spend their first week on campus finding 121 objects that would help them activate the Ruun, a sort of supercomputer capable of ascertaining the location of the Parasite. But there was no PS, no Parasite, no Ruun—at least not at first. It was all an elaborate fiction: The students were playing an alternate reality game.
Eos; Ali S. Akanda, Sonia Aziz, Antarpreet Jutla, Anwar Huq, Munirul Alam, Gias U. Ahsan, and Rita Colwell
from
A new initiative combines satellite data with ground observations to assess and predict the risk of cholera outbreaks in Bangladesh’s vulnerable populations.
It was hard to believe that Facebook would cozy up to disreputable advertisers in mid-2017 as it was under intense scrutiny from lawmakers and the media over revelations that Russian trolls had used the platform to influence the 2016 presidential election. Officially, the Berlin conference was for aboveboard marketing, but the attendees I spoke to dropped that pretense after the mildest questioning. Some even walked around wearing hats that said “farmin’,” promoting a service that sells fake Facebook accounts.
Granted anonymity, affiliates were happy to detail their tricks. They told me that Facebook had revolutionized scamming. The company built tools with its trove of user data that made it the go-to platform for big brands. Affiliates hijacked them. Facebook’s targeting algorithm is so powerful, they said, they don’t need to identify suckers themselves—Facebook does it automatically. And they boasted that Russia’s dezinformatsiya agents were using tactics their community had pioneered.
Delays in the testing and integration of NASA’s next space observatory, the James Webb Space Telescope (JWST), will push its launch back to May 2020, the agency announced today. NASA’s acting administrator, Robert Lightfoot, also admitted in a press briefing that the project’s cost may exceed the ceiling of $8 billion imposed by Congress in 2011. The agency expects to provide a confirmed schedule and cost estimate this summer. Congress will have to give its approval for extra spending if the cost cap has been breached.
The two parts of the spacecraft—the telescope and instrument package and the spacecraft bus with sunshield—are waiting to be melded together at the facility of prime contractor Northrop Grumman in Redondo Beach, California. The JWST was originally planned to launch in October, but last September NASA pushed back the launch to the second quarter of 2019.
The risk of dying from heart disease varies dramatically from one ZIP code to the next. Researchers are teasing apart the reasons why.
Heart disease has been the leading cause of death in the United States for more than 100 years, and the fact that it is particularly devastating in certain parts of the country is not news. For decades, people in the southern and southeastern states have died from cardiovascular disease at higher rates than those in the rest of the country. But when geographic details are examined with a finer lens, the disparities become even more pronounced—and sometimes less explicable.
“Literally from one highway exit to the next, you can see dramatic differences in mortality from heart disease,” says Gregory Roth, assistant professor at the Institute for Health Metrics and Evaluation (IHME) at the University of Washington School of Medicine in Seattle. He and his IHME collaborators have been looking at this variation for some time, and they recently published a series of maps in JAMA showing it in county-by-county detail.
Extra Extra
Nicholas Bloom at Stanford ran a study that shows working from home leads to a big burst of productivity for workers with long commutes (workers with short commutes were not offered the option to work from home). It is unclear if the boost would remain over time, and he is now quite the cheerleader for working from home. What I found in my dissertation research on this question: 1) after about 2-6 months, workers get stir crazy and want an office; 2) there are fewer interruptions at home so certain types of work are best at home (quiet reading, intensive writing, many kinds of programming); and 3) working from home all the time is bad for collaboration in a variety of ways. My clever subjects were able to replace some of what they lost in proximity with co-temporality – being in the same time or rhythm – on chat channels, skype/hangouts; even listening to the same streaming music feed helped. If anyone wants to partner up on a study that tests the hypotheses that an 80% office, 20% home strategy is ideal for productivity, creativity and satisfaction, let me know.
Charlottesville, VA April 12. “The second annual Applied Machine Learning Conference will convene researchers, entrepreneurs, and practitioners who use big data and machine learning applications for a day of presentations and flash talks.” [$$]
Our research has shown that researchers, HEIs, academic journals and funding bodies are yet to develop consistent guidelines and approaches towards the ethical use of social media Data. To that end, and accepting that the use of social media data for research purposes is likely to grow in prominence, we offer the following guidelines.