Computers Math

When you search for something on the internet, do you scroll through page after page of suggestions — or pick from the first few choices?
Because most people choose from the tops of these lists, they rarely see the vast majority of the options, creating a potential for bias in everything from hiring to media exposure to e-commerce.
In a new paper, Cornell University researchers introduce a tool they’ve developed to improve the fairness of online rankings without sacrificing their usefulness or relevance.
“If you could examine all your choices equally and then decide what to pick, that may be considered ideal. But since we can’t do that, rankings become a crucial interface to navigate these choices,” said computer science doctoral student Ashudeep Singh, co-first author of “Controlling Fairness and Bias in Dynamic Learning-to-Rank,” which won the Best Paper Award at the Association for Computing Machinery SIGIR Conference on Research and Development in Information Retrieval.
“For example, many YouTubers will post videos of the same recipe, but some of them get seen way more than others, even though they might be very similar,” Singh said. “And this happens because of the way search results are presented to us. We generally go down the ranking linearly and our attention drops off fast.”
The researchers’ method, called FairCo, gives roughly equal exposure to equally relevant choices and avoids preferential treatment for items that are already high on the list. This can correct the unfairness inherent in existing algorithms, which can exacerbate inequality and political polarization, and curtail personal choice.
“What ranking systems do is they allocate exposure. So how do we make sure that everybody receives their fair share of exposure?” said Thorsten Joachims, professor of computer science and information science, and the paper’s senior author. “What constitutes fairness is probably very different in, say, an e-commerce system and a system that ranks resumes for a job opening. We came up with computational tools that let you specify fairness criteria, as well as the algorithm that will provably enforce them.”
Algorithms seek the most relevant items to searchers, but because the vast majority of people choose one of the first few items in a list, small differences in relevance can lead to huge discrepancies in exposure. For example, if 51% of the readers of a news publication prefer opinion pieces that skew conservative, and 49% prefer essays that are more liberal, all of the top stories highlighted on the home page could conceivably lean conservative, according to the paper.
“When small differences in relevance lead to one side being amplified, that often causes polarization, where some people tend to dominate the conversation and other opinions get dropped without their fair share of attention,” Joachims said. “You might want to use it in an e-commerce system to make sure that if you’re producing a product that 30% of people like, you’re getting a certain amount of exposure based on that. Or if you have a resume database, you could formulate safeguards to make sure it’s not discriminating by race or gender.”
The research was partly supported by the National Science Foundation and by Workday.

Story Source:
Materials provided by Cornell University. Original written by Melanie Lefkowitz. Note: Content may be edited for style and length. More

New research led by Flinders University PhD candidate Jasmine Petersen examining commercial physical activity apps has found that the social components of these apps hold great potential to increase physical activity engagement.
Sharing physical activity outcomes and progress to app communities and social networking platforms provides the necessary encouragement for people to engage more enthusiastically with their apps.
“Sharing posts and receiving encouragement provides the social support many people need to stay motivated with exercise programs — and this doesn’t change across different age groups,” says study co-author Dr Ivanka Prichard, from Flinders University’s Caring Futures Institute.
The study — “Psychological mechanisms underlying the relationship between commercial physical activity app use and physical activity engagement,” by Jasmine Petersen, Lucy Lewis, Eva Kemps and Ivanka Prichard — is published in Psychology of Sport and Exercise.
The study examined close to 1300 adults (88% female, aged between 18 and 83 years), over half of whom used a commercial physical activity app (e.g. Fitbit, Garmin, Strava). Results found that more competitive individuals responded best to the apps, engaging in significantly higher levels of physical activity due to the game-like incentives and rewards built into the apps.
Dr Prichard says this suggests that people with a general disposition toward competition may benefit most from using activity apps.
“App users are motivated by both the enjoyment derived from physical activity (intrinsic motivation) and the personal value placed on the outcomes of physical activity (identified regulation), and these combined motivations result in greater engagement in physical activity,” says Ms Petersen.
This study shows that the social components of physical activity apps are particularly beneficial in promoting engagement in physical activity due to their capacity to facilitate social support, and positively influence motivation and beliefs in one’s ability to perform physical activity.
However, it was also found that online interactions can have a negative effect on exercisers if social networking is used to make direct comparisons.
“Engagement in comparisons was associated with lower self-efficacy and higher external regulation, and in turn, lower physical activity,” says Dr Prichard, emphasising the importance of exercising for enjoyment and the benefits that exercise can provide to general health.
The team are now following up participants to see how commercial physical activity apps might support physical activity behaviour in light of COVID-19 restrictions.

Story Source:
Materials provided by Flinders University. Original written by Megan Andrews. Note: Content may be edited for style and length. More

Simply put, string theory is a proposed method of explaining everything. Actually, there’s nothing simple about it. String theory is a theoretical framework from physics that describes one-dimensional, vibrating fibrous objects called “strings,” which propagate through space and interact with each other. Piece by piece, energetic minds are discovering and deciphering fundamental strings of the physical universe using mathematical models. Among these intrepid explorers are Utah State University mathematicians Thomas Hill and his faculty mentor, Andreas Malmendier.
With colleague Adrian Clingher of the University of Missouri-St. Louis, the team published findings about two branches of string theory in the paper, “The Duality Between F-theory and the Heterotic String in D=8 with Two Wilson Lines,” in the August 7, 2020 online edition of ‘Letters in Mathematical Physics.’ The USU researchers’ work is supported by a grant from the Simons Foundation.
“We studied a special family of K3 surfaces — compact, connected complex surfaces of dimension 2 — which are important geometric tools for understanding symmetries of physical theories,” says Hill, who graduated from USU’s Honors Program with a bachelor’s degree in mathematics in 2018 and completed a master’s degree in mathematics this past spring. “In this case, we were examining a string duality between F-theory and heterotic string theory in eight dimensions.”
Hill says the team proved the K3 surfaces they investigated admit four unique ways to slice the surfaces as Jacobian elliptic fibrations, formations of torus-shaped fibers. The researchers constructed explicit equations for each of these fibrations.
“An important part of this research involves identifying certain geometric building blocks, called ‘divisors,’ within each K3 surface,” he says. “Using these divisors, crucial geometric information is then encoded in an abstract graph.”
This process, Hill says, enables researchers to investigate symmetries of underlying physical theories demonstrated by the graph.
“You can think of this family of surfaces as a loaf of bread and each fibration as a ‘slice’ of that loaf,” says Malmendier, associate professor in USU’s Department of Mathematics and Statistics. “By examining the sequence of slices, we can visualize, and better understand, the entire loaf.”
The undertaking described in the paper, he says, represents hours of painstaking “paper and pencil” work to prove theorems of each of the four fibrations, followed by pushing each theorem through difficult algebraic formulas.
“For the latter part of this process, we used Maple Software and the specialized Differential Geometry Package developed at USU, which streamlined our computational efforts,” Malmendier says.

Story Source:
Materials provided by Utah State University. Original written by Mary-Ann Muffoletto. Note: Content may be edited for style and length. More

Jonathan Boreyko, an associate professor in mechanical engineering, has developed an aircraft thermal management technology that stands ready for adaptation into other areas.
The research was published in Advanced Functional Materials on Aug. 18, 2020.
Boreyko was the recipient of a Young Investigator Research Program award in 2016, given by the Air Force Office of Scientific Research. This award funded the development of planar bridging-droplet thermal diodes, a novel approach to thermal management. Boreyko’s research has shown this new approach to be both highly efficient and extremely versatile.
“We are hopeful that the one-way heat transfer of our bridging-droplet diode will enable the smart thermal management of electronics, aircraft, and spacecraft,” said Boreyko.
Diodes are a special kind of device that allow heat to conduct in only one direction by use of engineered materials. For management of heat, diodes are attractive because they enable the dumping of heat entering one side, while resisting heat on the opposite side. In the case of aircraft (the focus of Boreyko’s funding), heat is absorbed from an overheated plane, but resisted from the outside environment.
Boreyko’s team created a diode using two copper plates in a sealed environment, separated by a microscopic gap. The first plate is engineered with a wick structure to hold water, while the opposite plate is coated with a water-repelling (hydrophobic) layer. The water on the wicking surface receives heat, causing evaporation into steam. As the steam moves across the narrow gap, it cools and condenses into dew droplets on the hydrophobic side. These dew droplets grow large enough to “bridge” the gap and get sucked back into the wick, starting the process again.
If the source of heat were instead applied the hydrophobic side, no steam can be produced because the water remains trapped in the wick. This is why the device can only conduct heat in one direction.
What does this look like in practice? An object producing heat, like a CPU chip, overheats if this heat is not continually removed. Boreyko’s invention is affixed to this heat source. Generated heat is transferred through the conducting plate, into the water. Water turns to steam and moves away from the source of the heat. The hydrophobic, nonconducting side prevents heat from entering via the air or other heat sources that may be near, allowing the diode to manage the heat only from its main subject.
Boreyko’s team measured a nearly 100-fold increase in heat conduction when the wicked side was heated, compared to the hydrophobic side. This is a significant improvement to existing thermal diodes. According to Boreyko, current diodes are either not very effective, only conducting a few times more heat in one direction, or require gravity. This new bridging-droplet thermal diode can be used upright, sideways, or even upside-down, and would even work in space where gravity is negligible.
The team has filed a provisional patent and is in search of industry partners to carry on the work.

Story Source:
Materials provided by Virginia Tech. Note: Content may be edited for style and length. More

Patients who need surgery to fix major bone fractures use fewer opioid pills after their procedure if they’re reminded of their values — and those reminders don’t necessarily need to come from a doctor, according to a new study published in the Journal of Medical Internet Research.
“We showed that opioid medication utilization could be decreased by more than a third in an at-risk patient population by delivering psychotherapy via a chatbot,” said the study’s lead author, Christopher Anthony, MD, the associate director of Hip Preservation at Penn Medicine and an assistant professor of Orthopaedic Surgery. “While it must be tested with future investigations, we believe our findings are likely transferrable to other patient populations.”
Although opioids can be appropriate to treat the pain that results from an injury like a broken leg or arm, there is a concern that a large prescription of opioids might be an on-ramp to dependence for many. The researchers — who included Edward Octavio Rojas, MD, a resident in Orthopaedic Surgery at the University of Iowa Hospitals & Clinics — believe a low-effort, patient-centered approach to reducing the number of opioids taken can be a valuable method for cutting into the opioid epidemic.
To test this approach, 76 patients who went to a Level 1 Trauma Center at the University of Iowa Hospitals & Clinics for fractures that required a surgery to fix were randomly divided into two groups. Although each group received the same prescription of an opioid medication for pain, just one group was enrolled in a daily text-messaging program. That group received two daily text messages to their phones for two weeks after their procedure from an automated “chatbot” — a computer that uses artificial intelligence to send messages — starting the day after their surgery. The goal of each message was to help focus patients and hone their coping skills for the inevitable pain after such a procedure.
While they don’t expressly discourage using opioid pills, the messages, designed by a pain psychologist who specialized in acceptance and commitment therapy (ACT), are designed to direct thoughts away from taking a painkiller.
Each message fell under one of six “core principles”: Values, Acceptance, Present Moment Awareness, Self-As-Context, Committed Action, and Diffusion.
So, for example, a message a patient could receive under the Acceptance principle could be: “Feelings of pain and feelings about your experience of pain are normal after surgery. Acknowledge and accept these feelings as part of the recovery process. Remember how you feel now is temporary and your healing process will continue. Call to mind pleasant feelings or thoughts that you experienced today.” Or a Committed Action message might urge a patient to work toward a life goal, even if some pain might be present.
Overall, the patients who didn’t receive the messages took 41 opioid tablets after their surgeries, on average. The group who were regularly contacted by the chatbot averaged just 26, a 37 percent difference. Moreover, they reported less pain, overall, just two weeks after their procedure.
Importantly, the messages each patient received were not curated for their individual personality. This type of effectiveness was seen without the messages needing to be overly personalized. Combined with the using a chatbot instead of a human-intensive effort, this could be a low-cost, low-effort for orthopaedic and other procedures that provides significant protection from opioid dependence.
“A realistic goal for this type of work is to decrease opioid utilization to as few tablets as possible, with the ultimate goal being to eliminate the need for opioid medication in the setting of fracture care,” Anthony said.
This study was funded by a grant from the Orthopaedic Trauma Association.
Co-authors included Valerie Keffala, PhD; Natalie Ann Glass, PhD; Benjamin J. Miller, MD; Mathew Hogue, MD; Michael Wiley, MD; Matthew Karam, MD; and John Lawrence Marsh, MD, all of the University of Iowa, as well as Apurva Shah, MD, of the Children’s Hospital of Philadelphia. More

COMPUTER SCIENCE Researchers from the University of Copenhagen and the Technical University of Denmark (DTU) thought that they were five years away from solving a math riddle from the 1980’s. In reality, and without knowing, they had nearly cracked the problem and had just given away much of the solution in a research article. The solution could be used to improve tomorrow’s phones and computers.
Jacob Holm and Eva Rotenberg
The two computer scientists, Assistant Professor Jacob Holm of UCPH and Associate Professor Eva Rotenberg of DTU almost gave their solution away in the summer of 2019, after submitting a research article that became the precursor to the article in which they finally solved the math riddle.
A veritable brain teaser. That’s how one can safely describe this mathematical problem in the discipline of graph theory. Two mathematicians from the University of Copenhagen’s Department of Computer Science and DTU have now solved a problem that the world’s quickest and most clever have been struggling with since the 1980’s.
The two computer scientists, Assistant Professor Jacob Holm of UCPH and Associate Professor Eva Rotenberg of DTU almost gave their solution away in the summer of 2019, after submitting a research article that became the precursor to the article in which they finally solved the math riddle.
“We had nearly given up on getting the last piece and solving the riddle. We thought we had a minor result, one that was interesting, but in no way solved the problem. We guessed that there would be another five years of work, at best, before we would be able to solve the puzzle,” explains Jacob Holm, who is a part of BARC, the algorithm section at UCPH’s Department of Computer Science.

advertisement

The three utilities problem
In 1913, a precursor to the now solved mathematical conundrum was published in “The Strand Magazine” as “The Three Utilities Problem”. It caused the magazine’s readers to scratch their heads and ponder. In the problem, each of three cottages must have water, gas and electricity, while the “lines” between the houses and water, electricity and gas may not cross each other — which is not possible.
A solution between the lines
Simply put, the puzzle is about how to connect a number of points in a graph without allowing the lines connecting them to cross. And how, with a mathematical calculation — an algorithm — you can make changes to an extensive “graph network” to ensure that no lines intersect without having to start all over again. Properties that can be used for, among other things, building immense road networks or the tiny innards of computers, where electrical circuitry on circuit boards may not cross.
Jacob Holm has been interested in the mathematical conundrum since 1998, but the answer was only revealed while the two researchers were reading through their already submitted research article. In the meantime, the researchers heard about a novel mathematical technique that they realized could be applied to the problem.

advertisement

“While reading our research article, we suddenly realized that the solution was before our eyes. Our next reaction was ‘oh no – we’ve shot ourselves in the foot and given away the solution,’ says Associate Professor Eva Rotenberg of DTU.
About graph theory
A GRAPH is a very simple construction used to model things that can be described as objects and the connections between them. Graph theory is both an area of mathematics and an important tool in computer science.
In this context, a graph can be illustrated by a diagram consisting of a number of points (nodes, vertices) associated with a number of lines (edges). Each edge is illustrated as a line (or curved piece) with nodes as its two endpoints.
About the solution
There are two kinds of updates in dynamic graphs: One can delete an edge and you can insert a new edge. These two operations must be made by the user, while an algorithm keeps track of the network’s drawing at all times. This is the algorithm that the researchers have found the recipe for.
Read the research article: https://arxiv.org/abs/1911.03449
Could be used for computer electronics
This is when the two researchers got busy writing the research paper and tying up loose ends to solve the conundrum that Holm had been working on intermittently since 1998.
“We worked on the article non-stop, for five to six weeks. And, it ended up filling more than 80 pages,” says Eva Rotenberg.
Fortunately, no one beat them to the solution and the two researchers were able to present their results at the main theoretical computer science conferences, which were meant to be held in Chicago, but ended up being held virtually.
So, what can the solution to this mathematical conundrum be used for? The two researchers don’t know for sure, but they have a few suggestions.
“Our research is basic research, so we rarely know what it will end up being used for. Even from the start, we find applications difficult to imagine,” says Jacob Holm, who adds:
“the design of microchips and circuit boards, found in all electronics, could be an area where our result ends up being used. When drawing wires on a circuit board, they must never intersect. Otherwise, short circuits will occur. The same applies to microchips, which contain millions of transistors and for which one must have a graph drawing.”
### More

Like electronic devices, biological cells send and receive messages, but they communicate through very different mechanisms. Now, scientists report progress on tiny communication networks that overcome this language barrier, allowing electronics to eavesdrop on cells and alter their behavior — and vice versa. These systems could enable applications including a wearable device that could diagnose and treat a bacterial infection or a capsule that could be swallowed to track blood sugar and make insulin when needed.
The researchers will present their results today at the American Chemical Society (ACS) Fall 2020 Virtual Meeting & Expo.
“We want to expand electronic information processing to include biology,” says principal investigator William E. Bentley, Ph.D. “Our goal is to incorporate biological cells in the computational decision-making process.”
The new technology Bentley’s team developed relies on redox mediators, which move electrons around cells. These small molecules carry out cellular activities by accepting or giving up electrons through reduction or oxidation reactions. Because they can also exchange electrons with electrodes, thereby producing a current, redox mediators can bridge the gap between hardware and living tissue. In ongoing work, the team, which includes co-principal investigator Gregory F. Payne, Ph.D., is developing interfaces to enable this information exchange, opening the way for electronic control of cellular behavior, as well as cellular feedback that could operate electronics.
“In one project that we are reporting on at the meeting, we engineered cells to receive electronically generated information and transmit it as molecular cues,” says Eric VanArsdale, a graduate student in Bentley’s lab at the University of Maryland, who is presenting the latest results at the meeting. The cells were designed to detect and respond to hydrogen peroxide. When placed near a charged electrode that generated this redox mediator, the cells produced a corresponding amount of a quorum sensing molecule that bacteria use to signal to each other and modulate behavior by altering gene expression.
In another recent project, the team engineered two types of cells to receive molecular information from the pathogenic bacteria Pseudomonas aeruginosa and convert it into an electronic signal for diagnostic and other applications. One group of cells produced the amino acid tyrosine, and another group made tyrosinase, which converts tyrosine into a molecule called L-DOPA. The cells were engineered so this redox mediator would be produced only if the bacteria released both a quorum sensing molecule and a toxin associated with a virulent stage of P. aeruginosa growth. The size of the resulting current generated by L-DOPA indicated the amount of bacteria and toxin present in a sample. If used in a blood test, the technique could reveal an infection and also gauge its severity. Because this information would be in electronic form, it could be wirelessly transmitted to a doctor’s office and a patient’s cell phone to inform them about the infection, Bentley says. “Ultimately, we could engineer it so that a wearable device would be triggered to give the patient a therapeutic after an infection is detected.”
The researchers envision eventually integrating the communication networks into autonomous systems in the body. For instance, a diabetes patient could swallow a capsule containing cells that monitor blood sugar. The device would store this blood sugar data and periodically send it to a cell phone, which would interpret the data and send back an electronic signal directing other cells in the capsule to make insulin as needed. As a step toward this goal, VanArsdale developed a biological analog of computer memory that uses the natural pigment melanin to store information and direct cellular signaling.
In other work, Bentley’s team and collaborators including Reza Ghodssi, Ph.D., recently designed a system to monitor conditions inside industrial bioreactors that hold thousands of gallons of cell culture for drug production. Currently, manufacturers track oxygen levels, which are vital to cells’ productivity, with a single probe in the side of each vessel. That probe can’t confirm conditions are uniform everywhere in the bioreactor, so the researchers developed “smart marbles” that will circulate throughout the vessel measuring oxygen. The marbles transmit data via Bluetooth to a cell phone that could adjust operating conditions. In the future, these smart marbles could serve as a communication interface to detect chemical signals within a bioreactor, send that information to a computer, and then transmit electronic signals to direct the behavior of engineered cells in the bioreactor. The team is working with instrument makers interested in commercializing the design, which could be adapted for environmental monitoring and other uses.

Story Source:
Materials provided by American Chemical Society. Note: Content may be edited for style and length. More

Although true “cyborgs” — part human, part robotic beings — are science fiction, researchers are taking steps toward integrating electronics with the body. Such devices could monitor for tumor development or stand in for damaged tissues. But connecting electronics directly to human tissues in the body is a huge challenge. Now, a team is reporting new coatings for components that could help them more easily fit into this environment.
The researchers will present their results today at the American Chemical Society (ACS) Fall 2020 Virtual Meeting & Expo. 
“We got the idea for this project because we were trying to interface rigid, inorganic microelectrodes with the brain, but brains are made out of organic, salty, live materials,” says David Martin, Ph.D., who led the study. “It wasn’t working well, so we thought there must be a better way.”
Traditional microelectronic materials, such as silicon, gold, stainless steel and iridium, cause scarring when implanted. For applications in muscle or brain tissue, electrical signals need to flow for them to operate properly, but scars interrupt this activity. The researchers reasoned that a coating could help.
“We started looking at organic electronic materials like conjugated polymers that were being used in non-biological devices,” says Martin, who is at the University of Delaware. “We found a chemically stable example that was sold commercially as an antistatic coating for electronic displays.” After testing, the researchers found that the polymer had the properties necessary for interfacing hardware and human tissue.
“These conjugated polymers are electrically active, but they are also ionically active,” Martin says. “Counter ions give them the charge they need so when they are in operation, both electrons and ions are moving around.” The polymer, known as poly(3,4-ethylenedioxythiophene) or PEDOT, dramatically improved the performance of medical implants by lowering their impedance two to three orders of magnitude, thus increasing signal quality and battery lifetime in patients.
Martin has since determined how to specialize the polymer, putting different functional groups on PEDOT. Adding a carboxylic acid, aldehyde or maleimide substituent to the ethylenedioxythiophene (EDOT) monomer gives the researchers the versatility to create polymers with a variety of functions.
“The maleimide is particularly powerful because we can do click chemistry substitutions to make functionalized polymers and biopolymers,” Martin says. Mixing unsubstituted monomer with the maleimide-substituted version results in a material with many locations where the team can attach peptides, antibodies or DNA. “Name your favorite biomolecule, and you can in principle make a PEDOT film that has whatever biofunctional group you might be interested in,” he says.
Most recently, Martin’s group created a PEDOT film with an antibody for vascular endothelial growth factor (VEGF) attached. VEGF stimulates blood vessel growth after injury, and tumors hijack this protein to increase their blood supply. The polymer that the team developed could act as a sensor to detect overexpression of VEGF and thus early stages of disease, among other potential applications.
Other functionalized polymers have neurotransmitters on them, and these films could help sense or treat brain or nervous system disorders. So far, the team has made a polymer with dopamine, which plays a role in addictive behaviors, as well as dopamine-functionalized variants of the EDOT monomer. Martin says these biological-synthetic hybrid materials might someday be useful in merging artificial intelligence with the human brain.
Ultimately, Martin says, his dream is to be able to tailor how these materials deposit on a surface and then to put them in tissue in a living organism. “The ability to do the polymerization in a controlled way inside a living organism would be fascinating.” More