Dear Computer Science Alumni and Friends,
This past June, I shook hands with nearly 400 graduates as they crossed the stage at Frost Amphitheater to receive their diplomas. Commencement is one of the few times I wear heels of any significant height—it's an attempt to close the gap slightly between my stature and that of your average graduate—so my feet were killing me. On top of that, my mortarboard slipped a few millimeters back with each handshake. But I couldn't have been happier! It’s the first time we've held the computer science commencement ceremony in Frost's beautiful and majestic setting. Our traditional venue has been the courtyard of the Gates Computer Science building, but we've outgrown that space: Over the past five years, the number of undergraduate CS majors has almost quadrupled. I mentioned in last year's newsletter that CS is the largest major on campus; now we're almosttwice the size of the second-largest major, human biology. Later in the newsletter I'll mention some of our secrets to achieving such unprecedented popularity. (Incidentally, those 400 graduates were comprised of roughly 220 BS degrees, 150 MS degrees, and 30 PhDs.)
Another number that continues to rise, to our great delight, is the percentage of our undergraduate majors who are women. While the national average lingers around 15%, we're dramatically higher at 25%, and growing. We owe a debt of gratitude to some of our upperclass women CS majors, who have tirelessly made themselves available to chat with underclass women, and to organize events where prospective majors can meet amazing women alums who are making a splash in their jobs and loving it. I make a point of checking in with women CS majors about how they’re feeling in a traditionally male-dominated field. More and more often, they don’t even know why I’m asking. We’re not done yet, but that’s great progress.
This past year also saw the approval of a new Joint Major Program in CS with the humanities, enabling Stanford's multi-faceted students to explore and integrate two very different areas of interest. We hired two promising young faculty and one well-known senior professor, continued to conduct high-impact research across the entire field, and garnered the usual slew of student, faculty, and staff awards. All of these developments are covered below. Finally, as an example of our outreach activities, I’ll describe an inspiring program that brought introductory computer science to low-income high school students in Turkey.
On a personal note, after 5½ years I'm finally stepping down as CS department chair. It's been an extraordinary experience to serve the department, and to learn more about the School of Engineering and the university. I liken our department to a large ship moving through the ocean. The role of the department chair is largely to ensure that the ship plows smoothly through the waves, and doesn't get blown off course. My experience has been that the CS ship is so stable and so strong, only a light touch on the helm has been needed, and it's been an honor.
In the remainder of this newsletter, please enjoy the many departmental highlights of the 2013-14 academic year.
We’re exceedingly pleased to have three new faculty joining the department in the fall of 2014. While on the surface they represent well-established subfields of computer science—networking, artificial intelligence, and human-computer interaction—each of them boasts an unusual portfolio of application areas and technical interests. Keep an eye on these guys!
· Stefano Ermon is receiving his Ph.D. from Cornell University under the supervision of Profs. Carla Gomes and Bart Selman. Stefano works broadly in the area of Artificial Intelligence, with research contributions across probabilistic reasoning, machine learning, and control theory. Stefano has applied many of his results and techniques to a new area known as computational sustainability, with breakthroughs ranging from fishery management to optimizing electric-car battery use. Recognizing his leadership in this emerging field, Stefano will also be named a Fellow of Stanford's Woods Institute for the Environment. We're very excited about this new connection.
· James Landay comes to us after holding faculty positions at UC Berkeley, University of Washington, and Cornell NYC Tech. We're confident that Stanford will be James' final resting place! James fills a critical gap in the department's Human-Computer Interaction group, which had shrunk to just one junior faculty member at the same time the HCI field is burgeoning in popularity. With his vast leadership experience, a history of building vibrant research groups and attracting top graduate students, and a strong interdisciplinary ethic, we expect James to be a force in the department while forging bridges with the d.school and across the university.
· Keith Winstein is receiving his Ph.D. from MIT under the supervision of Prof. Hari Balaksrishnan. Keith works in the area of computer networks and systems. He has developed several widely used innovative software systems, including the Sprout transport protocol for high throughput in wireless networks, the Mosh remote terminal application for mobile clients, and Remy, a program that uses Bayesian inference to generate congestion-control algorithms. In his "other life," Keith spent three years as a reporter for the Wall Street Journal, and he's long been involved in issues surrounding copyright law and internet policy. Keith will hold a courtesy faculty appointment in Stanford's Law School, another connection we're very excited about.
Tenure and Retirement
I'm pleased to announce that Prof. Gill Bejerano, jointly appointed in CS and Developmental Biology, has been promoted to Associate Professor with tenure. Gill’s work at the intersection of two fields, combining laboratory benches with serious computation, is a wonderful example of the type of interdisciplinary research we expect to become more and more prevalent in the CS department and across the university.
Meanwhile, Prof. Marc Levoy has retired his regular faculty position, although we look forward to Marc’s continued contributions to departmental teaching and student advising. Marc created many exciting and highly-visible research projects during his time at Stanford. His work on 3D laser scanning in the 1990s culminated in the Digital Michelangelo Project, in which he and his students spent a year in Italy digitizing statues. His 3D scanning techniques are now widely used to make special effects for movies. Marc is also well-known for his work on computational photography, a now ubiquitous term he first coined in 2004 for a Stanford course. In 1996, together with Prof. Pat Hanrahan, Marc invented light field imaging, which allows the capture of real-world objects in a way that provides the ability to change focus and perspective after the imagery is captured. This technology is now known to consumers through the Lytro light field camera.
Post-retirement, Marc is working at Google, where he has been involved in many projects including the Google book scanner, Street View, and cameras for Google Glass and Android devices.
Popularity of the CS Major: Our Secret Sauce
As department chair, these days I'm frequently what our trick is—how we've attracted so many students to major in computer science. I've been assured it's not that the admissions office has decided to favor techies. In fact, a large fraction of our majors had no intention of pursuing computer science, or any other form of engineering for that matter, when they arrived at Stanford. So what happened? Soon enough, every freshman hears about a legendary class: CS106A. Perhaps peer pressure is a contributor, but one way or another a huge fraction of Stanford students are now braving the CS106A waters, some with considerable initial trepidation. We have truly outstanding instructors in the CS106 courses, who quickly and ably convey to a wide variety of students the challenges and joy of creating computer programs, and of computer science as a field of study. The students soon realize that it can be a lot more fun to learn a subject where everyone finds his or her own unique solution to each problem, rather than being in a class where getting a different answer from one's classmates means getting it wrong. As students move on to subsequent CS courses, they learn that computer science is a unique combination of engineering, design, intuition, logic, creativity, and perseverance, and they're hooked!
New CS+X Joint Major Program
Many Stanford students have strong interests in fields in addition to computer science, yet they choose computer science as their sole major because double-majoring is notoriously difficult (necessitating all requirements of two majors, with no double-counting of classes), and a double major provides neither time nor a structured opportunity for integrating the two fields. Students who are interested in both a humanities field, say, and CS, are more likely to choose CS as their major for purely practical reasons. What a shame that these students aren't able to pursue both of their passions! Enter the new Joint Major Program, dubbed "CS+X".
Starting next year, the university has approved CS+X programs for ten majors across seven partner humanities departments. The intention is to eventually include partners in the social sciences, natural sciences, and other parts of the university, but the pilot is of limited scope to get things started and see how they go. While music and linguistics are obvious partners, we're also excited about the prospect of joint majors in English, philosophy, and Slavic languages, among others.
On the computer science side, students undertaking the Joint Major may drop two electives from the CS major requirements, and their senior project is required to be a capstone that integrates CS and the partner major. Computation and data are rapidly becoming integral to a wide variety of other fields, so students and their advisors should have little difficulty coming up with interesting projects. We have a long tradition of CS department research collaborations outside of the department. We're very excited to now offer interdisciplinary opportunities to eager undergraduates.
Introductory Computer Science for Turkish High-Schoolers
K-12 education in Turkey has a strong tradition in math and science, but the country has a relatively small presence in software. In June 2014, a team from the CS department taught a two-week intensive class in Java programming to high school students at Darussafaka boarding school in Istanbul. The students at Darussafaka are from low-income families across Turkey, and all of them have lost one or both parents. In just two weeks, 25 eager 10th and 11th graders covered the entire CS106A curriculum, from Karel to Breakout and beyond. The class was organized by Prof. Nick McKeown and his Turkish wife Asena; the teaching team included some of our very talented TAs and section leaders: Chris Piech, Bryce Cronkite-Ratcliff, Julia Lee, Nick Troccoli, and TY Huang.
As always, a tremendous amount of exciting research is going on in the department. Please enjoy these three samples, and visit the CS Department website to learn about our ongoing research across the entire field.
We frequently complain of poor mobile wireless connectivity. Yet, paradoxically, we are awash with connectivity options. Walking around a city, we see tens of WiFi access points and several cellular options, but our devices cannot access the available capacity. The OpenRadio project, led byProf. Sachin Katti, virtualizes wireless networks, enabling users to stitch together the connectivity they need from the networks around them. Instead of a fixed bill every month to access a single network, users pick and choose the best options at any point depending on their needs, availability, and cost. In the long term, OpenRadio decouples wireless network service from the infrastructure, and enables the definition of wireless radio networks in software, from the lowest radio level to higher layer network services. Software definition enables more flexible networks, efficient use of scarce spectrum, and in the long run turns wireless physical infrastructure (such as cell towers) into a shared utility on top of which different service provides can define their network service in software. Then, networks will compete on the quality, richness, and flexibility of their services, and not on who owns more cell towers as is the case today.
Crowdsourcing a Web Site in One Day
What could you accomplish if you could summon together software developers, designers, and user researchers, all on demand and within minutes? Could you transform a napkin sketch of a design idea into a working, user-tested mobile application in 24 hours? Prof. Michael Bernsteinhas been working with a group of students to create computationally-managed flash teams of crowdsourced experts. Their platform has managed flash teams from online labor markets (e.g., oDesk) to create web applications, short animations from scratch, and an online class platform complete with content. Experiments suggest the flash teams are twice as efficient as self-managed teams. Underlying flash tasks is an online system for authoring teams and managing them at runtime through structured handoffs. This system can grow and shrink teams on demand, pipeline output, and combine modular team elements to create entire small organizations.
Carbon Nanotube Computer
Further miniaturization of computing systems in silicon is limited fundamentally by physics. Representing a radical departure from traditional silicon, transistors with semiconducting channels made of carbon nanotubes offer a significant opportunity to solve this outstanding challenge. After the first demonstration of carbon-nanotube transistors in 1998, there was great excitement about a new age of energy-efficient computing systems. However, challenging material imperfections inherent to carbon nanotubes stalled progress. Recently, Prof. Subhasish Mitra (joint with EE), together with his EE colleague Prof. Philip Wong and their students, made a major breakthrough, making it possible to build robust computing systems that perform correctly despite substantial carbon nanotube imperfections. Their paper describing the first carbon-nanotube based microprocessor was featured on the cover of Nature. Prof. Mitra and Wong's widely-publicized milestone promises to transform the landscape of exploratory nanotechnologies from stand-alone transistor demonstrations to actual nanosystems.
Faculty and Staff Awards and Honors
It's been another banner year for major faculty awards:
· Dan Boneh ― ACM Goedel Prize, Fellow of International Association for Cryptologic Research
· Ed Feigenbaum ― George R. Stibitz Computer & Communications Pioneer Award
· Pat Hanrahan ― Technical Achievement Academy Award, Katayanagi Prize
· Sachin Katti ― ACM SIGCOMM Rising Star
· Oussama Khatib ― IEEE Robotics and Automation Society Distinguished Service and Georges Saridis Awards
· Anshul Kundaje ― Sloan Foundation Fellow
· Percy Liang – Microsoft Faculty Fellow
· John Ousterhout ― IEEE Reynold B. Johnson Information Storage Systems Award
· Balaji Prabhakar ― IEEE Innovation in Societal Infrastructure Award
· Mendel Rosenblum ― American Academy of Arts and Sciences, IEEE Reynold B. Johnson Information Storage Systems Award
· Tim Roughgarden ― 2014 Social Choice and Welfare Prize
· Mehran Sahami ― ACM Presidential Award
In addition, our beloved and hard-working student services officer, Claire Stager, won the Shah Award for School of Engineering Staff Member of the Year.
Our students’ outstanding work was recognized with several awards in research, teaching, and programming this year.
· Research Awards ― Amy Sentis won this year’s Wegbreit Prize for Undergraduate Computer Science Research for her honors thesis titled Discriminating Brain Activation Patterns for Cognitive and Sensory Aspects of Pain Processing. The Arthur Samuel Award for outstanding CS Ph.D. thesis went to Aditya Parameswaran for his dissertation titled Human Powered Data Management. The Christofer Stephenson Award for best MS research report was given to co-winners Nikhil Khadke (Snapworld: A Self-Tuning Distributed Graph Processing Framework) and Ming Han Teh (DataSift: An Expressive and Accurate Crowd-Powered Search Toolkit).
· Teaching Awards ― The CS Department Forsythe Teaching Award went to co-winners Justin Solomon and Michael Chang. Justin has not only worked as an outstanding TA, but he has taught three courses, and has written a textbook for one of them! Michael began as a CS106 section leader his freshman year, and has gone on to section lead and TA on a regular basis, most recently serving as head TA for CS107 (Computer Organization & Systems). Michael, who also won a School of Engineering Centennial TA Award, is the instructor for CS107 this summer. Firas Abuzaid was also selected for a Centennial TA Award. Like Michael, Firas began as a CS106 section leader, growing into a highly sought-after head TA for large courses, and this summer is co-teaching CS145 (Introduction to Databases).
· ACM International Collegiate Programming Contest ― Stanford put in another solid showing at the ACM ICPC, with our top two teams taking 2nd and 3rd place out of 117 teams at the Pacific Northwest Regional Contest. First place went to Berkeley. Stanford’s top team, along with the Berkeley team, advanced to the ACM ICPC World Finals in Russia. Not to be outdone by our cross-bay rival a second time, the team of Jack Chen, Joshua Wang, and Hieu Pham ranked ahead of Berkeley at the World Finals. Nice work!
Until the next newsletter, have a terrific year.
Fletcher Jones Professor
Chair, Department of Computer Science