2014年7月14日 - 2014年7月15日

Faculty Summit 2014

地点： Redmond, WA, USA

Monday, July 14

  • Chair: Mary L. Gray, Microsoft Research

    Join us for a conversation to reflect on the ethics, implications, and responsibilities of social media research, in the wake of the Facebook emotion study. What obligations must researchers consider when studying human interaction online? When does data science become human subjects research? What can we learn as a collective from the public’s reaction to Facebook’s recent research as well as reflection on our own work? Mary L. Gray (Microsoft Research) and Jeff Hancock (Cornell University and co-author of the Facebook emotion study), will facilitate a panel discussion among researchers based at Microsoft Research and across academia from the fields of data science, computational social science, qualitative social science, and computer science.

  • Speakers: Hadi Esmailzadeh, Georgia Institute of Technology; Chris Harrison, Carnegie Mellon University; Brandon Lucia, Microsoft Research; Kris Pister, University of California-Berkeley

    Over the years, we have been successful making computing devices smaller, more efficient, and cheaper. Unfortunately, barriers such as constraints on power and energy mean we can no longer trivially shrink today’s devices any further. In this session, we will explore creative approaches to new challenges in user interaction, low-power system designs, novel form factors, and hardware innovations.

  • Speakers: Elizabeth Gerber, Northwestern University; Michael Luca, Harvard Business School; Jeff Nickerson, Stevens Institute of Technology; Arun Sundararajan, New York University

    From Airbnb, to Kickstarter, to Mechanical Turk, the on-demand provisioning of workforce, services, and goods is changing the nature of work and play. In this session, we focus on exploring the challenges and opportunities with these platforms that span from the technological to the social. We will also discuss the future of crowd-based work. The session will feature a wide range of speakers involved in crowdsourcing research as well as lively discussion panels.

  • Speakers: Jeff Bilmes, University of Washington; Jianfeng Gao and Xiaodong He, Microsoft Research; Mari Ostendorf, University of Washington; Paul Smolensky, Johns Hopkins University; Eric Xing, Carnegie Mellon University

    Deep learning has enjoyed tremendous success in recent years in speech and visual object recognition, as well as in language processing (although to somewhat less extent). The focus of this session is on deep learning approaches to problems in language or text processing, with particular emphasis on important applications with vital significance to Microsoft. First, we will have both academic and Microsoft Research experts provide a tutorial on the latest deep learning technology, presenting both theoretical and practical perspectives on common methods of deep neural networks and recurrent, recursive, stacking, and convolutional networks. We will highlight special challenges faced by language/text processing, and elaborate on how new deep learning technologies are poised to fundamentally address these issues. We will share Microsoft Research’s experience in developing Deep-Structured Semantic Models (DSSM) and their successful applications to web search, ads selection, machine translation, and entity search.

  • Speakers: Jennifer Breslin, UN Women; Kelly Cox, Iron Ways Film; Karen Peterson, National Girls Collaborative Project; Lucy Sanders, National Center for Women and Information Technology

    Join us on a journey. Big Dream, a series and film produced by Microsoft Research and Iron Way Films, centers around the stories of five young women around the globe working on transforming the world through technology. Each story will show an intimate look at these young women’s ambitions, challenges, and innovative approaches to their lives. You will also hear voices of women— both accomplished and starting their computer science career—who share their fears, aspirations, advice, and their big audacious dreams. The Big Dream Campaign will launch in the fall of 2014. You will get an early glimpse into three of the five stories. You will meet the producer, researchers, and students in the film and have the opportunity to ask them questions. You will learn what role your university and you can take to join the Big Dream Movement. Join us to be inspired and see how we can grow more women in computing.

  • Speakers: Nilanjan Bannerjee, University of Maryland; Arjmand Samuel, Microsoft Research; Affan Syed, FAST-NUCES, Pakistan

    An increasing number of research areas rely on collecting data from sensors and devices deployed in homes and beyond. Researchers typically deploy such devices, collect data, analyze and make interesting inferences based on this data. In order to collecting sufficient data to have confidence in the research findings it is desirable to collect data from a large numbers of locations. However, doing so requires major investment in engineering expertise and technology infrastructure; both not readily available to the academic community. Microsoft Research’s Lab of Things aims to provide such an infrastructure to facilitate at-scale in-situ research in a number of research areas. In this session, academic researchers deploying studies using the Lab of Things will provide overview of their in-situ research and lessons learnt while deploying such studies.

  • Speakers: Jeff Bigham, Carnegie Mellon University; Eric Horvitz and Ben Livshits, Microsoft Research; Haoqi Zhang, Northwestern University–Evanston

    Abstract is the same as Part I:

    From Airbnb, to Kickstarter, to Mechanical Turk, the on-demand provisioning of workforce, services, and goods is changing the nature of work and play. In this session, we focus on exploring the challenges and opportunities with these platforms that span from the technological to the social. We will also discuss the future of crowd-based work. The session will feature a wide range of speakers involved in crowdsourcing research as well as lively discussion panels.

  • Speakers: Misha Bilenko, Microsoft; Rich Caruana, Microsoft Research; Isabelle Guyon, Chalearn

    This session will look into the latest advances in areas of machine learning, such as causality, while also reviewing our understanding of topics such as deep learning, and how to scale machine learning.

    This session will also highlight steps towards doing reproducible science by enabling researchers to share code and data, and experiments to help nurture an environment of scientific rigor. And it will open up new avenues for collaboration between researchers via the use of co-opetitions where people can cooperate with each other to reach a higher value than by merely competing.

    Misha Bilenko will talk about “Scaling Up Machine Learning: A Production Perspective,” Rich Caruana will address the question of “Do Deep Nets Really Need to Be Deep?”, and Isabelle Guyon present the latest “Contribution of Machine Learning Challenges to Causal Discovery.”

  • Speakers: Jonathan Carlson, Microsoft Research; Lee Hood, Institute for Systems Biology; Eli Van Allen, Harvard University; Ravi Pandya, Microsoft Research

    A hardware revolution has slashed the cost of sequencing by over a million-fold in a few years, far faster than Moore’s law, opening great opportunities for software to revolutionize healthcare. This session will explore how the combination of the $1,000 genome with molecular diagnostics and electronic medical records, across millions of patients, will transform medicine, making it not just personalized, but also predictive, preventive, and participatory. You will hear about a pioneering program to follow 100,000 well patients, building for each individual a multi-dimensional data model to optimize their wellness and minimize their disease. We will also discuss how genomic data is being used in the clinic to personalize cancer therapies to individual patients;  the computational analysis of HIV evolution, transmission, and immune response, to improve the effectiveness of therapeutics and vaccines; and how to efficiently scale genomic analysis into the cloud.

  • Speaker: Leslie Lamport, Turing Award winner, Microsoft Research

    Architects draw detailed blueprints before a brick is laid or a nail is hammered. Programmers and software engineers seldom do. A blueprint for software is called a specification. The need for extremely rigorous specifications before coding complex or critical systems should be obvious—especially for concurrent and distributed systems. This talk explains why some sort of specification should be written for any software.

Tuesday, July 15

  • Speaker: Keith Hampton, Associate Professor, Rutgers University

    Social media, we have been promised, supports the well-being of individuals and society. It has been said that it offers new opportunities for democratic participation, that it supports collective action and gives individuals a voice. Social capital, once lost through the dormancy of ties, hidden as a result of infrequent contact, has been made visible through the persistent and pervasive nature of social media. Yet, it is often said that these technologies do not live up to their hype, that they contribute to social isolation, stress, and disengagement. This presentation argues that community on and offline is being reorganized as a result of digital technologies that afford and constrain engagement. Evidence from a series of empirical studies explains how digital technologies are related to social capital, deliberation, caring, and the joy that we sometimes get from the awareness that we are all missing out.

  • Speakers: John Carroll, Pennsylvania State University; Ed Cutrell, Microsoft Research; Andrea Kavanaugh, Virginia Polytechnic Institute & State University; Robert Mason, University of Washington

    Communication technologies have long played a central role in people’s personal and professional lives. However, as social media has become globally ubiquitous in the past decade, enabling people to connect with any person, friend or stranger, around the world, communication technologies have also played an increasingly important role in the public, civic sphere. These networked publics provide a new form of third place—a place outside the home and work—transforming how people are connected and mobilized to address civic issues. Concurrently, progressive governments are redefining societal prosperity to encompass not only financial success but also the quality of life of its citizens. An important condition of well-being is membership in a thriving community that effectively collaborates both to have fun and to solve its collective problems. In this panel, we will discuss how we should help people leverage networked publics as third places to positively impact global citizen well-being.

  • Speakers: Rakesh Agrawal, Sumit Basu, and Anoop Gupta, Microsoft Research

    Availability of high quality education is widely acknowledged to be the pathway to success in modern society. The past few years have seen a tremendous interest in use of MOOCs, SPOCs, flipped-classrooms / blended-learning to provide more scalable and affordable models for student learning. However, it is still hard to author interactive online lessons, so only a small fraction of faculty create or use them. This session will introduce Office Mix, a brand new offering from Microsoft that dramatically simplifies the creation of such online lessons, including their publishing and sharing, and associated analytics. Office Mix builds upon the deep familiarity of faculty and students with PowerPoint to create such lessons, and use the slide decks they already have in their arsenal. We will also discuss use cases beyond online learning, to sharing and communication of academic research.

    The session will also cover two other efforts from Microsoft Research. Sumit Basu will show Powergrading, a powerful method for increasing the efficiency of grading students answers to online short-answer questions. Rakesh Agarwal will discuss technologies for inferring a knowledge graph from current education material, enriching the graph with rich content in multiple format mined from the web as well as crowd-sourcing, and then overlaying it with the social graph of teachers and students to enable dynamic formation of study teams with the goal of maximizing overall learning.

  • Speakers: Jeff Bilmes, University of Washington; Chris Manning, Stanford University; Nima Mesgarani, Columbia University; Dong Yu, Microsoft Research

    The last few years have witnessed a renaissance in multiple areas of speech and language processing. In speech recognition, deep neural networks have led to significant performance improvements; in language processing the idea of continuous-space representations of words and language has become mainstream; and dialog systems have advanced to the point where automated personal assistants are now everyday fare on mobile devices. In this session, we bring together researchers from the different disciplines of speech and language processing to discuss the key ideas that have made this possible, and the remaining challenges and next generation of applications.

  • Speakers: Michael Freedman, Princeton University; Bhuvan Urgaonkar, Pennsylvania State University; Adam Wierman, California Institute of Technology

    Online apps that matter are geo-distributed, because they serve millions of users around the world. Geo-distribution is important for latency, availability, and increasingly also for efficiency. Due to rapid growth in the volume of demand served, large numbers of geo-distributed data centers today can benefit from the same multi-megawatt economies of scale that were initially limited to a few centralized ones. As a result, modern cloud infrastructures are already highly geo-distributed. Trends such as our increasing reliance on online services, and the growth in mobile device usage that has converted many client based applications into cloud services, indicate that data center infrastructures will be even more geo-distributed in the future. This session discusses challenges in designing a geo-distributed data center infrastructure and developing software applications for it.

  • In a world with a billion sensors, how will we make sense of it all? In our daily lives we encounter sensors all the time, like when a motion sensor turns a light on in a dark place, or when a carbon monoxide detector tell us that the air is becoming hazardous. Sensors extend our abilities to see, hear, and feel far beyond what we ourselves can take in—from arrays of telescopes sensing the edges of the universe to nano-scale biological sensors amplifying our own sense of smell.

    In a world with a billion sensors, how will we make sense of it all?

    How will sensors change the way we perceive not only our environment but ourselves and others?

    How will sensors change the way we live and work?

    What interfaces, services, devices and experiences will be necessary to make sense of it all and avoid sensory overload?

    What are key problems this data can be used to help solve, what new troubles can we anticipate it creates?

    • Carnegie Mellon University (opens in new tab), Department of Design, Pittsburgh, PA, United States
      • Professors: Peter Scupelli, Bruce Hanington
      • Microsoft Liaisons: Jane Park, Annika Ushio
      • Project: Grassroots, a platform for neighbors to connect by creating and sharing data from the neighborhood’s sensor network
    • Royal Danish Academy of Fine Arts, Product Design, Copenhagen, Denmark
      • Professor: Troels Degn Johansson
      • Microsoft Liaison: Jakob Nielsen
      • Project: Navi-Band: A tool toy for safety, navigation and play.
    • Goldsmiths (opens in new tab), University of London, United Kingdom
      • Professor: Tobie Kerridge
      • Microsoft Liaisons: Richard Banks, Tim Regan
      • Project: Wired Eye: A kaleidoscope for viewing aerial shots of data centers
    • Art Center College of Design, Pasadena, CA, United States
      • Professors: Phil Van Ellen, Ben Hooker
      • Microsoft Liasons: Colleen Estrada, Joyce Chou, Malek Chalabi
      • Project: The Future of Wearable Services: A Proposal for a Pop-Up Sensor Nail Salon

  • Speakers: Rajesh Patel, Microsoft; Matt Lease, University of Texas at Austin; Siddharth Suri, Microsoft Research

    This presentation draws together experts in the field to discuss the technical and social challenges posed by crowdsourcing’s global workflows. While crowdwork is often presumed to be driven by anonymous, autonomous actors, this panel draws on a diverse set of research approaches to flesh out a more socially robust vision of this vital part of the information economy. We examine crowdsourcing’s systems of exchange and collaboration as well as the range of networks and incentives that organize crowdwork. Each panelist offers a perspective from current findings, from computational and qualitative approaches, to ask: who are crowdworkers and how might seeing who they are help us build more responsive, expansive, and ethical platforms?

  • Speakers: Hyunju Lee, Gwangju Institute of Science and Tecnology; Parker MacCready, University of Washington-Seattle; Yan Xu, Beihang University; Chunmiao Zheng, Peking Univeristy

    The sciences are currently undergoing a fundamental transition due to the avalanche of data that is generated by instruments, simulations, on-line archives and social media. The impact of the data revolution is seen in every discipline. Cloud computing was invented to manage the big data challenges of Internet companies, but it is now seen as a critical tools for many research communities. Life Science, environmental science and geosciences have been early adopters of cloud technology because of easy by which the cloud can accrue data from many sources and make it available for analysis by large communities. This session illustrates work by academic researchers who have been awarded “Microsoft Azure for Research” cloud awards. We highlight four of our early project out of the 190 projects that we have awarded. We will also briefly discuss some new tools for machine learning and data analysis in the cloud that we will be making available to the community.

  • Speakers: Grey Ballard, Sandia National Laboratories; Madan Musuvathi, Microsoft Research; Keshav Pingali, University of Texas at Austin

    Parallelism abounds in modern hardware—from the datacenter to multi-cores, GPUs, and FPGAs. On the other hand, important algorithms, such as graph algorithms, dynamic programming, and finite-state machine processing involve fine-grained dependencies and do not directly map on to this parallel hardware. Harnessing the parallelism available for these algorithms requires new algorithms, new programming languages, and new runtime systems. This session will present recent advances in this area and will serve as forum for bringing together researchers from diverse disciplines—such as algorithms, programming languages, compilers and runtime, machine-learning, architecture, and systems—into this exciting research area.

  • Speakers: David Reilly, University of Sydney; Rob Schoelkopf, Yale University; Dave Wecker, Microsoft

    Three decades have passed since Richard Feynman first proposed devising a “quantum computer” founded on the laws of quantum physics to achieve computational speed-ups over classical methods. In that time, quantum algorithms have been developed that offer fast solutions to problems in a variety of fields including number theory, chemistry, and materials science. To execute such algorithms on a quantum device will require extensive quantum and classical “software”. One of the grand challenges for the computer science community is the design and implementation of a software architecture to control and program quantum hardware. This session will address how to build a scalable, reliable quantum computer: What are the quantum and classical resource requirements? How do we protect the device against errors? How do we program the quantum computer? It will highlight recent advances in quantum device architectures, error correction, and software design tools, and pose crucial open questions in quantum computer science.

  • Speakers: Adrian Caulfield, Microsoft Research; Martha Kim, Columbia University; Tom Wenisch, University of Michigan

    The slowing and eventual ending of Moore’s Law will dramatically impact datacenter operators, who have long depended on steady advances in server performance and efficiency to make improved services economically viable. Specialization in the form of hardware accelerators (e.g., FPGAs, GPGPUs, and ASICs) can overcome performance and energy limitations but introduce challenging problems at scale, such as cost, flexibility, programmability, and the need to gracefully integrate with existing software stacks. This session will invite experts within Microsoft Research and the external research community to discuss challenges and opportunities for specialization in the cloud.

  • Speakers: Andrew Begel and Mary Czerwinski, Microsoft Research; Erin Solovey, Drexel University

    Confused and frustrated by your difficult programming task? Let’s have your IDE intervene before you get the chance to cause a bug in the software. Bored by your rote assignment? Let’s have your browser find a funny YouTube video to snap you out of it. Stressed out? Let’s Snapchat your friends so they can call and lend you an ear. What else could a computer do if it knew what you were feeling? In this session, we’ll talk about several research projects which leverage low-cost biometric sensors to sense your affect and take action to help you out.