The latest issue of Research Library Issues (RLI) looks at the critical role and participation of libraries and librarians in supporting the data science revolution at research universities. … This issue of RLI asserts that research libraries are prepared for the data science revolution as we draw upon our long-standing contributions: creating the conditions for new knowledge discovery, teaching students how to discern validity, and partnering with the research community to prepare and preserve data for science.
Interactions.org, Lawrence Berkeley National Laboratory
from
It took three sky surveys – conducted at telescopes in two continents, covering one-third of the visible sky, and requiring almost 1,000 observing nights – to prepare for a new project that will create the largest 3D map of the universe’s galaxies and glean new insights about the universe’s accelerating expansion.
This Dark Energy Spectroscopic Instrument (DESI) project will explore this expansion, driven by a mysterious property known as dark energy, in great detail. It could also make unexpected discoveries during its five-year mission.
The surveys, which wrapped up in March, have amassed images of more than 1 billion galaxies and are essential in selecting celestial objects to target with DESI, now under construction in Arizona.
One of the rationales behind a corporate investment in IoT is that it provides the data needed to understand products and equipment so well that they will never go down. Because once a company can predict the behavior of a product or machine, it can guarantee that product or machine’s performance. And from that point, it can offer that product or machine as a service.
That is a big technological and business shift, however. On the tech side, getting the sensors in place and setting up reliable connectivity is only the first step. After that, a company has to take the data coming from the machines those sensors are embedded into and figure out how to derive insights about product health from it. Which is tougher than it might sound. Companies can go through dozens of algorithms before finding one that works, and sometimes they have to adjust their expectations. For example, maybe they can’t predict exactly when a machine will fail, but they can predict when it will most need a tune-up.
But once the tech is in place — including the sensors, the connectivity, and the appropriate algorithm — there are business challenges to contend with. The biggest one is that many customers aren’t keen to buy a product as a service. Peter Zornio, CTO of Emerson Automation Solutions, explains that Emerson has been offering some of its products as a service for a few years, but only a small percentage of customers have signed up.
The chart above represents sample travel times over the past 14 months, based on more than a year of trip records from the Metropolitan Transportation Authority. Each dot represents one commute, including waits and transfers but not walks to and from stations.
It reveals a wide and often frustrating variability in morning subway commutes — something most statistics fail to capture, yet something most New Yorkers intuitively understand.
It’s all about incentives. The current academic ecosystem incentivises publication in high impact factor journals and grant capture above all else, but there is more to being an academic than producing journal articles and winning grants. Luckily there are an increasing number of initiatives that are helping academics get credit for more of the work they do and increase their broader impact. This post rounds up some of the most interesting efforts.
Facial recognition technology can be used to prevent criminal activity. But, in London, one study shows the police system gets it wrong 81 percent of the time.
This week, public records revealed how ICE has scanned facial data from millions of driver’s license photos. But the agency’s collaboration with state DMVs might go deeper.
In a step toward molecular storage systems that could hold vast amounts of data in tiny spaces, Brown University researchers have shown it’s possible to store image files in solutions of common biological small molecules.
The Vera Institute of Justice announced The University of Scranton Center for the Analysis and Prevention of Crime was among the 16 organizations in the nation to be awarded an In Our Backyards Community Grant to provide data analysis and public information around the need to reduce incarceration in Lackawanna County.
As part of this project, analysts from the University will set up a data-sharing, analysis and dissemination plan with the Lackawanna County Prison and other community partners.
“Incarceration is just one way society can deal with offending,” said Michael Jenkins, Ph.D., executive director of the Center for the Analysis and Prevention of Crime. “Unfortunately, our system has relied too much on prison as a crime prevention tool. We’ve learned a lot about the devastating social and economic effects of incarceration, and, as a result, people from all different backgrounds and ideologies recognize the need to reduce our reliance on it.”
If you’re in health care and don’t live under a rock, you have probably heard that United Health Group (UHG) has acquired PatientsLikeMe (PLM). After the announcement, there was a lot of sound and fury, some of which signified nothing, as the saying goes, and some which signified a lot.
Three good friends – Susannah Fox, Jane Sarasohn-Kahn and Lisa Suennen – got to talking about this and realized we had so much to say we just had to write it down in one giant melting pot of prose – a trifecta of thoughts about this transaction but, more globally, about the entire burgeoning phenomenon of data as a business and patients as…People? An asset to be sold or bartered? A sum of their data? We hope it’s the former, but we also worry it’s the latter two at times.
Get a cup of coffee and sit back, because this is a long one – not the usual short and sweet entry that each of us endeavors to craft on our respective blogs.
The chairman of the Federal Trade Commission asked children’s privacy advocates whether having video creators on YouTube disable ads could resolve concerns the site is violating laws to protect kids, according to a person familiar with the conversation.
During a July 1 call, Chairman Joseph Simons and fellow Republican Commissioner Noah Phillips suggested the world’s largest video site wouldn’t need to move all children’s content to a separate platform as advocates have proposed, according to the person. Instead, individual channels could disable advertising to bring the site into line with a U.S. law’s ban on collecting information on children under age 13 without parental permission.
The Securities and Exchange Commission used a proprietary algorithm to spot suspicious trading that will soon send a former Goldman Sachs Group Inc. banker to prison.
That case is only one example of the rapid adoption of artificial intelligence across the U.S. government. About half of the top 100 regulatory agencies are now using one or more types of AI to carry out their daily work, according to researchers from Stanford and New York universities who are cataloging its use and expect to publish their findings later this year.
Perhaps because of the vast scale at which most social platforms operate, proposed solutions to the problem of online hate speech tend to be technical in nature. In theory, a platform might fine-tune its algorithms to deëmphasize hate speech and conspiracy theories. But, in practice, this is harder than it sounds. Some overtly hateful users may employ language and symbols that clearly violate a site’s community guidelines—but so called borderline content, which dances at the edge of provocation, is harder to detect and draws a broad audience. Machine-learning systems struggle to tell the difference between actual hate speech and content that describes or contests it. (After YouTube announced its new policies, the Southern Poverty Law Center complained that one of its videos, which was meant to document hate speech, had been taken down.) Some automated systems use metadata—information about how often a user posts, or about the number of comments that a post gets in a short period of time—to flag toxic content without trying to interpret it. But this sort of analysis is limited by the way that content bounces between platforms, obscuring the full range of interactions it has provoked.
Tech companies have hired thousands of human moderators to make nuanced decisions about speech. YouTube also relies on anonymous outside “raters” to evaluate videos and help train its recommendations systems. But the flood of questionable posts is overwhelming, and sifting through it can take a psychological toll.
Nuclear power plants generate 20% of the nation’s electricity and are the largest clean energy source in the U.S. But to further offset climate change, the nuclear energy sector needs to extend the lifetime of existing facilities as well as build new ones.
This requires the U.S. switching from traditional analog technology to the latest advances in digital technology, a change already made in other countries.
The U.S. Nuclear Regulatory Commission has licensed Purdue University Reactor Number One (PUR-1) as the first entirely digital nuclear reactor instrumentation and control system in the nation. The upgraded reactor and facility, originally built in 1962, paves the way for widespread implementation of digital technology in both research and industry reactors.
Instagram launched a new feature today, Restrict, intended to help vulnerable users avoid abuse. Facebook’s Head of Instagram Adam Mosseri says the company will also be focusing on new uses for AI to crack down on bullying.
Social media platforms are under pressure by governments to show they can police themselves in matters that impact elections. In the spirit of showing they are responsible and don’t need a heavy regulatory hand, Instagram has declared war on bullying, and there’s quite a press launch to kick it off this week.
Boston, MA Conference is September 7 at Boston University. “This year we are offering ten travel scholarships. Our conference scholarships bring us one step closer in our mission. As part of the application, you are to develop an article on an artificial intelligence concept on the Machine Intelligence Community platform. This can be a summary of a reseach paper, an article explaining an algorithm or math concept, or research that you are working on.” Deadline to apply is July 31.
“The National Humanities Center will offer up to 40 residential fellowships for advanced study in the humanities for the period September 2020 through May 2021. Applicants must have a doctorate or equivalent scholarly credentials. Mid-career and senior scholars are encouraged to apply.” Fellowship competition closes on October 10.
“In many languages, there are many ways to solve a given problem. Programmers can spend a lot of time thinking about the best way to solve it. Golang, on the other hand, believes in fewer features — with only one right way to solve the problem.”
This article is the first one in a series that will be dedicated to forming a path for channelling out deep learning projects in a holistic manner. We will start off by discussing the importance of having a good strategy to structure deep learning projects. We will then decompose the units that are responsible for developing a deep learning project at a production scale and study the first set of units.