Data Science newsletter – June 9, 2020

GROUP CURATION: N/A

Researchers expect long-term changes that reduce travel for work and conferences: part 6 in a series on science after the pandemic.

A Harvard study found that long-term exposure to microscopic soot in the air appears to be associated with higher death rates from the coronavirus.

But Trump’s EPA has recommended keeping the 2012 standards for microscopic soot that are linked to an estimated 45,000 deaths a year.

The plan overrides advice from EPA staff scientists.

IBM CEO Arvind Krishna called on Congress Monday to enact reforms to advance racial justice and combat systemic racism, while announcing the company was getting out of the facial recognition business.

The decision for IBM to get out of the facial recognition business comes amid criticism of the technology, employed by multiple companies, for exhibiting racial and gender bias.

Broadening Participation in Computing was the main focus of my research and service agenda for over a dozen years. I thought that the goal was to get more women and under-represented minorities into CS. CS is obviously valuable and important for students. I thought that we just had to help students with diverse backgrounds to realize that. Instead, the lack of diversity is the canary in the coal mine. I’m learning that our goal should be to change CS Education so that everyone is welcome and supported. CS is not a welcoming place. We CS teachers have structured our systems to keep people out, to limit access to that valuable and important knowledge. We spend so much time and energy on detecting cheating and on finding ways to limit access to our major, which sends the message that most people don’t belong. A better use of that time and energy might be to provide tutoring and change our curriculum so that more diverse students succeed. We need to send the message that we are willing to change in order to address historic and systemic inequities.

We have to change CS so that it serves the needs of our students and society. Using methods like Peer Instruction and curricula like Media Computation are steps in the right direction, since they are measurably better for women and underserved populations, but those are results from the few diverse students who even walk in our door — and too few CS teachers are even willing to adopt these small measures.

L’Institut quantique (IQ), a research organization located at the University of Sherbrooke in Québec, is joining the IBM Q Network as an IBM Q Hub. IQ will be the first Hub in Canada and the 14th globally.

The University of Sherbrooke has received $4.5 million in support from the Government of Québec to expand its quantum computing capacity as an IBM Q Hub. The funding will be used to secure cloud-based access to IBM’s quantum computing systems and software, including a 53-qubit system, currently the largest universal one available in the industry.

Hateful rhetoric is plaguing online discourse, fostering extreme societal movements and possibly giving rise to real-world violence. A potential solution to this growing global problem is citizen-generated counter speech where citizens actively engage in hate-filled conversations to attempt to restore civil non-polarized discourse. However, its actual effectiveness in curbing the spread of hatred is unknown and hard to quantify. One major obstacle to researching this question is a lack of large labeled data sets for training automated classifiers to identify counter speech. Here we made use of a unique situation in Germany where self-labeling groups engaged in organized online hate and counter speech. We used an ensemble learning algorithm which pairs a variety of paragraph embeddings with regularized logistic regression functions to classify both hate and counter speech in a corpus of millions of relevant tweets from these two groups. Our pipeline achieved macro F1 scores on out of sample balanced test sets ranging from 0.76 to 0.97—accuracy in line and even exceeding the state of the art. On thousands of tweets, we used crowdsourcing to verify that the judgments made by the classifier are in close alignment with human judgment. We then used the classifier to discover hate and counter speech in more than 135,000 fully-resolved Twitter conversations occurring from 2013 to 2018 and study their frequency and interaction. Altogether, our results highlight the potential of automated methods to evaluate the impact of coordinated counter speech in stabilizing conversations on social media.

The risk of transmission is complicated by, and intertwined with, the urgent moral stakes: Systemic racism suffuses the United States. The mortality gap between black and white people persists. People born in zip codes mere miles from one another might have life-expectancy gaps of 10 or even 20 years. Two racial inequities meet in this week’s protests: one, a pandemic in which black people are dying at nearly twice their proportion of the population, according to racial data compiled by the COVID Tracking Project at The Atlantic; and two, antiblack police brutality, with its long American history and intensifying militarization. Floyd, 46, survived COVID-19 in April, but was killed under the knee of a police officer in May.

Americans may wish the virus to be gone, but it is not. While the outbreak has eased in the Northeast, driving down the overall national numbers, cases have only plateaued in the rest of the country, and they appear to be on the rise in recent days in COVID Tracking Project data.

the “fog gaming” concept seems to be centered around converting Sega’s massive infrastructure of Japanese arcades and arcade machines into a kind of widely distributed streaming-gaming data center. Those cabinets—and the decently specc’ed CPUs and GPUs inside them—are only in active use by players for perhaps eight hours a day at a busy location, according to Adam Pratt, an arcade operator who runs industry website Arcade Heroes. The rest of the time, those machines could serve streaming gaming content to homebound players, without the need for an immense, Google Stadia-sized data center investment.

In theory, at least.

In a series of papers scheduled to be presented at the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Amazon researchers propose complementary AI algorithms that could form the foundation of an assistant that helps customers shop for clothes. One lets people fine-tune search queries by describing variations on a product image, while another suggests products that go with items a customer has already selected. Meanwhile, a third synthesizes an image of a model wearing clothes from different product pages to demonstrate how items work together as an outfit.

When seeking a causative factor for the onset of insulin resistance and low-grade inflammation that usually accompanies obesity and diabetes, in the past scientists have turned to bacterial lipopolysaccharide (LPS) as a triggering factor.

However, the causal role of gut microbiota in obesity and type 2 diabetes (T2D) remains elusive. While the vast majority of studies have focused on studying fecal microbiota in obese people with diabetes, little is known about the existence of microbial signatures in tissues and organs involved in metabolic control.

A new study by Anhê and colleagues uncovers a unique microbial profile in the liver, adipose tissue and plasma in people with morbid obesity and diabetes.

In the following sections, we outline the ways in which different types of mobile phone data can help to better target and design measures to contain and slow the spread of the COVID-19 pandemic. We identify the key reasons why this is not happening on a much broader scale, and we give recommendations on how to make mobile phone data work against the virus. [full text]

The pandemic has bored holes in budgets at universities and colleges across the US and beyond. To stem the spread of the coronavirus, higher education institutions shuttered their campuses, sent students home, and moved classes online. As a result, schools have lost millions of dollars in revenue—and many expect further losses due to the uncertainly about whether the 2020–21 school year can convene on campus this fall.

“All colleges and universities have been affected by the pandemic. Most are facing a serious cash flow crisis,” says Terry Hartle, the senior vice president of government and public affairs at the American Council on Education, an association of college presidents. “The vast majority of college and university presidents are worried about fall enrollment and the financial viability of their institutions.”

A lot is changing this year in higher education. Amid the ongoing pandemic caused by the infectious coronavirus disease 2019 (COVID-19), universities and graduate schools have had to adapt to entirely online instruction and have canceled fieldwork, closed labs, and faced declining revenues. These immediate changes have been forced upon programs by necessity, and they, along with negative impacts on many students from the current pandemic, will likely continue affecting higher education in the near future by, for example, decreasing application numbers. To bolster fall admissions, some graduate programs are temporarily dropping the Graduate Record Examinations (GRE) as an admissions requirement. However, dropping the GRE altogether, as a step toward equity and inclusivity in graduate admissions and education, has been a longer-term battle, with many terming it #GRExit on social media.

The National Science Foundation’s (NSF) Innovative High-Performance Computing program, has awarded $10 million to the National Center for Supercomputing Applications (NCSA) to deploy and operate Delta, an advanced computing and data resource that will shape the future of technology and practice in advanced research computing. With the deployment of the Delta system, NCSA takes an active role in expanding the adoption of GPU-based computing and continues to be at the forefront of advanced research computing.

Her case is one of a number that have experts questioning the utility of testing people over and over again after they’ve recovered. Experts have also raised doubts about the value of a growing mountain of studies that show some people test positive for weeks after infection. There’s concern that such tests are being misinterpreted to suggest people are infectious when they probably are not, and are keeping them from returning to work — or, in this particular case, being reunited with their loved ones, for no good reason.