MIT Research News' Journal
 
[Most Recent Entries] [Calendar View]

Thursday, March 9th, 2017

    Time Event
    12:50p
    Cutting down the clutter in online conversations

    From Reddit to Quora, discussion forums can be equal parts informative and daunting. We’ve all fallen down rabbit holes of lengthy threads that are impossible to sift through. Comments can be redundant, off-topic or even inaccurate, but all that content is ultimately still there for us to try and untangle.

    Sick of the clutter, a team from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) has developed “Wikum,” a system that helps users construct concise, expandable summaries that make it easier to navigate unruly discussions.

    “Right now, every forum member has to go through the same mental labor of squeezing out key points from long threads,” says MIT Professor David Karger, who was senior author on a new paper about Wikum. “If every reader could contribute that mental labor back into the discussion, it would save that time and energy for every future reader, making the conversation more useful for everyone.”

    The team tested Wikum against a Google document with tracked changes that aimed to mimic the collaborative editing structure of a wiki. They found that Wikum users completed reading much faster and recalled discussion points more accurately, and that editors made edits 40 percent faster.

    Karger wrote the new paper with PhD students Lea Verou and Amy Zhang, who was lead author. The team presented the work last week at ACM’s Conference on Computer-Supported Cooperative Work and Social Computing in Portland, Oregon.

    How it works

    While wikis can be a good way for people to summarize discussions, they aren’t ideal because users can’t see what’s already been summarized. This makes it difficult to break summarizing down into small steps that can be completed by individual users, because it requires that they spend a lot of energy figuring out what needs to happen next. Meanwhile, forums like Reddit let users “upvote” the best answers or comments, but lack contextual summaries that help readers get detailed overviews of discussions.

    Wikum bridges the gap between forums and wikis by letting users work in small doses to refine a discussion’s main points, and giving readers an overall “map” of the conversation.

    Readers can import discussions from places such as Disqus, a commenting platform used for publishers like The Atlantic. Then, once users create a summary, readers can examine the text and decide if they want to expand the topic to read more. The system uses color-coded “summary trees” that show topics at different levels of depth and lets readers jump between original comments and summaries.

    “Our aim is to harness collaborative summarization to save the reader from sifting through hours of unorganized forums to find the content they want,” says Zhang.

    Editors have a similar interface to readers, with options for marking comments as unimportant, summarizing and grouping comments and replies, and citing or quoting key parts of comments. The researchers used an automatic highlighting algorithm that picks out important sentences for editors, who are limited to 250-word summaries.

    "Discussion is a timeless medium—some of what we do today on Reddit looks like discussion on USENET in the early 1990s,” says Amy Bruckman, a professor at the Georgia Tech who was not involved with the paper. “Innovating in this space is hard, and that’s what makes Wikum impressive. They’ve made the first serious advance in helping people quickly understand long conversations. Their approach to recursive crowd sourced summarization has intriguing potential."

    In the future, the goal would be to have a diverse group of contributors organize and moderate discussions. The team plans to explore how to best motivate users to summarize content.

    The team is also collaborating with the Wikimedia Foundation to deploy Wikum with their editors and further improve the system. Specifically, they’re seeing how editors can use it to summarize large discussions about important changes to Wikipedia pages like Requests for Comment (RFCs), in order to more easily make final decisions.

    They also want to examine using Wikum in a chat structure to see if real-time summaries could improve digital messaging systems like the types used on apps such as Slack and Trello.

    “At a time of great enthusiasm for machine learning and data mining, it's worth remembering that human beings can be excellent processors of information if they are given the right tools to leverage their intelligence,” Karger says.

    2:00p
    Toward “valleytronic” devices for data storage or computer logic systems

    Faster, more efficient data storage and computer logic systems could be on the horizon thanks to a new way of tuning electronic energy levels in two-dimensional films of crystal, discovered by researchers at MIT.

    The discovery could ultimately pave the way for the development of so-called “valleytronic” devices, which harness the way electrons gather around two equal energy states, known as valleys.

    Engineers have for some time warned that we are reaching the limits of how small we can build conventional electronic transistors, which are based on electrons’ electrical charge.

    As a result, researchers have been investigating the utility of a property of electrons known as spin, to store and manipulate data; such technologies are known as spintronics.

    But as well as their charge and spin, electrons in materials also have another “degree of freedom,” known as the valley index. This is so-called because plotting the energy of electrons relative to their momentum results in a graph consisting of a curve with two valleys, which are populated by electrons as they move through a material.

    Harnessing this degree of freedom could allow information to be stored and processed in some materials by selectively populating the two valleys with electrons.

    However, developing such valleytronic devices requires a system to selectively control the electrons within the two valleys, which has so far proven very difficult to achieve.

    Now, in a paper published today in the journal Science, researchers led by Nuh Gedik, an associate professor of physics at MIT, describe a new way of using laser light to control the electrons in both valleys independently, within atomically thin crystals of tungsten disulfide.

    “The two valleys are exactly at the same energy level, which is not necessarily the best thing for applications because you want to be able to tune them, to change the energy slightly so that the electrons will move [from the higher] to the lower energy state,” Gedik says.

    Although this can be achieved by applying a magnetic field, even very powerful laboratory magnets with a strength of 10 tesla are only capable of shifting the valley energy level by around 2 millielectronvolts (meV).

    The researchers have previously shown that by directing an ultrafast laser pulse, tuned to a frequency very slightly below the resonance of the material, they were able to shift the energy of one of the valleys through an effect called the “optical Stark effect,” while leaving the other valley virtually unchanged. In this way they were able to achieve a shift in energy level of up to 20 meV.

    “The light and the electrons inside the material form a type of hybrid state, which helps to push the energy levels around,” Gedik says.

    In the latest experiment, the researchers discovered that by tuning the laser frequency to even further below resonance, and increasing its intensity, they were able to simultaneously shift the energy levels of both valleys and reveal a very rare physical phenomenon.

    While one valley still shifts as a result of the optical Stark shift as before, the other valley shifts through a different mechanism, known as the “Bloch-Siegert shift,” according to MIT physics PhD student Edbert Jarvis Sie, the paper’s lead author.

    Although the Bloch-Siegert shift was first predicted in 1940, and soon after helped inspire Willis Lamb to his 1955 Nobel Prize-winning discovery of the Lamb shift in hydrogen atoms, it has remained a considerable challenge to observe it experimentally in solids.

    Indeed, apart from so-called artificial atoms, the new mechanism has never been observed in solids until now, because the resulting shifts were too small, Sie says. The experiment performed at the Gedik Lab produced a Bloch-Siegert shift of 10 meV, which is 1,000 times larger than that seen previously.

    What’s more, the two effects — the Bloch-Siegert shift and optical Stark shift — have previously tended to take place in the same optical transition, meaning researchers have had difficulty disentangling the two mechanisms, Sie says.

    “In our work we can disentangle the two mechanisms very naturally, because while one valley exhibits the optical Stark shift, the other valley exhibits the Bloch-Siegert shift,” Sie says. “This can work so nicely in this material because the two mechanisms have a similar relationship with the two valleys. They are related by what is called time-reversal symmetry.”

    This should allow for enhanced control over valleytronic properties in two-dimensional materials, Nuh says. “It could give you more freedom in tuning the electronic valleys,” he says.

    The research team included Liang Fu, the Lawrence C. and Sarah W. Biedenharn Assistant Professor in MIT’s Department of Physics; Jing Kong, associate professor of electrical engineering at MIT; Chun Hung Lui, assistant professor of physics at the University of California at Riverside; and Yi-Hsien Lee, assistant professor at National Tsing-Hua University in Taiwan. The work was supported by the U.S. Department of Energy, the Gordon Betty Moore Foundation, the National Science Foundation, and the Ministry of Science and Technology of Taiwan.

    The paper is the first report of the Bloch-Siegert in a semiconductor, according to John Schaibley, an assistant professor of physics at the University of Arizona, who was not involved in the research.

    “Gedik and his colleagues show that they can control this energy shift in a three-atom thick semiconductor,” he says.  “By varying the polarization of their laser, they can use the Bloch-Siegert shift to control different electronic states.”

    2:00p
    Conquering metal fatigue

    Metal fatigue can lead to abrupt and sometimes catastrophic failures in parts that undergo repeated loading, or stress. It’s a major cause of failure in structural components of everything from aircraft and spacecraft to bridges and powerplants. As a result, such structures are typically built with wide safety margins that add to costs.

    Now, a team of researchers at MIT and in Japan and Germany has found a way to greatly reduce the effects of fatigue by incorporating a laminated nanostructure into the steel. The layered structuring gives the steel a kind of bone-like resilience, allowing it to deform without allowing the spread of microcracks that can lead to fatigue failure.

    The findings are described in a paper in the journal Science by C. Cem Tasan, the Thomas B. King Career Development Professor of Metallurgy at MIT; Meimei Wang, a postdoc in his group; and six others at Kyushu University in Japan and the Max Planck Institute in Germany.

    “Loads on structural components tend to be cyclic,” Tasan says. For example, an airplane goes through repeated pressurization changes during every flight, and components of many devices repeatedly expand and contract due to heating and cooling cycles. While such effects typically are far below the kinds of loads that would cause metals to change shape permanently or fail immediately, they can cause the formation of microcracks, which over repeated cycles of stress spread a bit further and wider, ultimately creating enough of a weak area that the whole piece can fracture suddenly.

    “A majority of unexpected failures [of structural metal parts] are due to fatigue,” Tasan says. For this reason, large safety factors are used in component design, leading to increased costs during production and component life.

    Tasan and his team were inspired by the way nature addresses the same kind of problem, making bones lightweight but very resistant to crack propagation. A major factor in bone’s fracture resistance is its hierarchical mechanical structure, so the team investigated microstructures that would mimic this in a metal alloy.

    The question was, he says, “Can we design a material with a microstructure that makes it most difficult for cracks to propagate, even if they nucleate?” Bone provided a clue to how to do that, through its hierarchical microstructure — that is, the way its internal structures have different patterns of voids and connections at many different length scales, with a lattice-like internal structure — that combines strength with light weight.

    The team developed a kind of steel that has three key characteristics, which combine to limit the spread of cracks that do form. Besides having a layered structure that tends to keep cracks from spreading beyond the layers where they start, the material has microstructural phases with different degrees of hardness, which complement each other, so when a crack starts to form, “every time it wants to propagate further, it needs to follow an energy-intensive path,” and the result is a great reduction in such spreading. Also, the material has a metastable composition; tiny areas within it are poised between different stable states, some more flexible than others, and their phase transitions can help absorb the energy of spreading cracks and even lead the cracks to close back up.

    To further understand the relative roles of these three characteristics, the team compared steels each with a combination of two out of the three key properties. None of these worked as well as the three-way combination, he says. “This showed us that our modification has better fatigue resistance than any of these.”

    The testing of such materials under realistic conditions is difficult to do, Tasan explains, partly because of “the extreme sensitivity of these materials to surface defects. If you scratch it, it’s going to fail much faster.” So meticulous preparation and inspection of test samples is essential.

    This finding is just a first step, Tasan says, and it remains to be seen what would be needed to scale up the material to quantities that could be commercialized, and what applications would benefit most. “Economics always comes into it,” he says. “I’m a metallurgist, and this is a new material that has interesting properties. Large industries such as automotive or aerospace are very careful about making changes in materials, as it brings extra effort and costs.”

    But there are likely to be several uses where the material would be a significant advantage. “For critical applications, [the benefits] are so critical that change is worth the extra trouble” about the cost, he says. “This is an alloy that would be more expensive than a basic low-carbon steel, but the property benefits have been shown to be quite exceptional, and it’s with much lower amounts of alloying metals (and hence, costs) than other proposed materials.”

    The research was supported by the European Research Council and MIT’s Department of Materials Science and Engineering. The team included Motomichi Koyama, Zhao Zhang, Kaneaki Tsuzaki, and Hiroshi Noguchi of Kyushu University in Fukuoka, Japan, and Dirk Ponge, and Dierk Raabe of the Max Planck Institute in Dusseldorf, Germany.

    << Previous Day 2017/03/09
    [Calendar]
    Next Day >>

MIT Research News   About LJ.Rossia.org