Andries van Dam, the newly appointed chair of the Computing Research Association Education Committee (CRA-E), is already hard at work getting the word out about the problems of computing education. He spoke to the Chronicle of Higher Education about the concerns and the future work necessary.
[The following guest post by CRA Chair Dan Reed originally appeared on Dan's blog, Reed's Ruminations. We're pleased to repost it here.]
Much has been written about declining enrollments in computer science, the image of computing among secondary school students, and the depressingly small numbers of women and minorities enrolled in computer science programs. There are many opinions about the root causes of our enrollment problems and at least as many opinions about possible solutions. The reality of the problem is not in dispute, however.
Slicing the Infinite Onion
As I reflect on the past thirty years of computer science curricula and my experience as both a student and a professor, I am often struck by how little has changed. The core elements of our curricula remain centered on formal languages and theory, data structures, programming languages and compilers, operating systems and computer architecture. These are the courses I took as an undergraduate in the 1970s, and we still teach their evolutionary variants today.
Around continuous and discrete mathematics, physical and biological science and this computing core, we have added successive layers to the computing curriculum onion: graphics and human-computer interaction, artificial intelligence, mobile and embedded devices, computational geometry, networks and distributed systems, numerical and scientific algorithms, parallel computing, databases and data mining, privacy and information security, just to name a few.
As this non-exhaustive list illustrates, the computing curriculum onion has grown ever larger and more complex, with each layer derived from new discoveries and technologies. I do not believe this expansion can continue indefinitely. Asymptotics do apply – the number of students will tend (indeed, is tending toward) zero as the knowledge and degree expectations approaches infinity. This must change.
Rethinking Computing Education
I believe we must rethink our computing education approaches in some deep and fundamental ways. First, as researchers and technologists we seek to reproduce students in our technical image, failing to acknowledge that most of our students will not develop compilers, write operating systems or design computer chips. Rather, they benefit from training in logical problem solving, knowledge of computing tools and their applicability to new domains.
In short, most of our graduates solve problems using computing rather than working in core computing technologies. We must recognize and embrace the universality of computing as a problem solving process and introduce computing via technically challenging and socially relevant problem domains.
The magic hierarchy of computing – from atoms to gates to bits to in-order instruction architecture and machine language to code translation to "hello world" was an attractive and emotionally enticing technology story to previous generations. It is often esoteric and off-putting to a generation of students reared on ubiquitous computing technology.
This does not mean we should eviscerate the intellectual core of computing. Rather, we must emphasize relevance and introduce computing as a means to solve problems. Show the importance of computing to elections and voting, energy management and eco-friendly design, health care and quality of life.
Second, we struggle to accept the fact that not every student needs detailed knowledge of every computing specialization. If I were to draw a tortured analogy with the history of automobile, drivers need not understand combustion dynamics, the stiff ODE solutions underlying antilock brakes or superheterodyne radio engineering. Drivers do need to understand how to operate a car safely and recognize the high-level principles underlying that operation.
All of this suggests we should create multiple educational tracks that emphasis the disparate aspects of computing, layered atop a smaller, common core. Of course, I could be wrong – I often am.
CRA-E Committee
To explore the future of computing education, CRA has chartered a new committee, CRA-E (E for education), chaired by Brown professor Andries (Andy) van Dam. The new committee seeks to understand how the broad computing community needs to move forward in order to develop principles and philosophy underlying the computing education of the future. As I noted in the press release:
I am delighted that Professor van Dam has agreed to service as the initial chair of CRA-E. Not only is Andy a distinguished and respected researcher, he is passionate about computing education, both its theory and its practice. Moreover, he has long worked to apply novel technologies to computing education.Andy will be assembling a committee to think deeply and strategically about the future of computing education, especially at the undergraduate level. I look forward to the outcome of these explorations.
The National Science Foundation (along with all other federal agencies) released its FY 09 Budget Request to Congress on Monday. We've already had some preliminary coverage of it, noting that, on the whole, computing research does pretty well. Late Monday afternoon NSF hosted a briefing on its budget to provide a little finer resolution look at what they hope to get from Congress in this appropriations season -- and we've got those details below (spoiler: they're pretty good).
But maybe just as importantly, NSF's Computer and Information Science and Engineering (CISE) directorate also provided some detail about how it plans to deal with the austere appropriation it received from Congress for FY 2008. Before we get to the relatively good news from the request, it's probably appropriate to close the book on the FY 2008 numbers. You'll recall that CISE had some big plans for FY 2008. We've listed some of the potential impacts on NSF overall from the omnibus funding levels in a previous post, but here's what we know specifically about CISE:
For FY 2009, NSF hopes to make up the ground lost in the omnibus by requesting significant increases for its research directorates. Overall, NSF would see its budget increase by 14 percent over FY 08, to $6.06 billion in FY 09. Within that increase, computing research is featured prominently in the request. The Foundation-wide, but CISE led, Cyber-enabled Discovery and Innovation program would expand considerably under the agency's plan, growing from $48 million in FY 08 to $100 million in FY 09, including $33 million in CISE. Additionally, the agency has proposed two new foundation-wide initiatives that have strong computing foci. The first is a $20 million investment in "Science and Engineering Beyond Moore's Law," which "aims to position the U.S. at the forefront of communications and computation capability beyond the physical and conceptional limitations of current systems." That program would be led by the Mathematics and Physical Sciences directorate, but CISE would control $6 million in awards. The second is a $15 million investment ($3.5 million in CISE) in "Adaptive Systems Technology" that focuses on "generating pathways and interfaces between human and physical systems that will revolutionize the development of novel adaptive systems."
Additionally, CISE would see its core research budget increase by 19.5 percent, or $104 million, in FY 09 under the President's plan -- essentially making up all the ground lost with the omnibus. Programs of note within the directorate include:
For now, it's enough to say that the budget appears to once again represent a good start for NSF and computing in the appropriations cycle. But it's just the start of a long, unpredictable process.
Next up, a focus on DOD IT R&D....
WASHINGTON, DC -- The Computing Research Association is pleased to announce the appointment of Andries van Dam, Brown University Professor of Technology and Education and Professor of Computer Science, to lead its new effort to improve the quality of computing education, particularly at the undergraduate level. Professor van Dam will chair CRA's new Education Committee, called CRA-E, charged to think broadly about the future of computing education
“I am delighted that Professor van Dam has agreed to service as the initial chair of CRA-E,” said Daniel Reed, Microsoft’s Scalable and Multicore Computing Strategist and Chair of CRA. “Not only is Andy a distinguished and respected researcher, he is passionate about computing education, both its theory and its practice. Moreover, he has long worked to apply novel technologies to computing education.”
CRA established the CRA-E after declining enrollments in computer science led to a reexamination of the image of computing and the nature of the 21st Century computing curriculum. The new committee seeks to understand how the broad computing community needs to move forward in order to develop principles and philosophy underlying the computing education of the future. “I don’t believe we can continue the indefinite addition of layers to the computing curriculum onion that was defined in the 1970s,” said Reed. “We need to rethink some of our fundamental assumptions about computing education approaches and content, and Professor van Dam is the right person to lead that effort.”
Professor van Dam has been on Brown’s faculty since 1965 and was one of the founders and first Chair of its Computer Science Department. Along with J.D. Foley, van Dam authored seminal texts on computer graphics, and has authored or co-authored over 100 papers. He is a fellow of the IEEE, Association for Computing Machinery, American Academy of Arts and Sciences, and the American Association for the Advancement of Science, a member of the National Academy of Engineering, recipient of awards for outstanding contributions to computing education from ACM, IEEE and SIGCSE, and a winner of the 2002 CRA Distinguished Service Award. van Dam also served as the Chair of CRA from 1985 to 1987.
About CRA. CRA was established 30 years ago and has members at more than 250 research entities in academia, industry and government. Its mission is to strengthen research and advance education in the computing fields, expand opportunities for women and minorities, and improve public and policymaker understanding of the importance of computing and computing research in society.
For more about CRA.
Congratulations to Edmund M. Clarke of Carnegie Mellon, E. Allen Emerson of UT-Austin, and Joseph Siafkis of Verimag Laboratory in France, on being awarded ACM's 2007 A.M. Turing award, the highest honor in computing, for their work on a quality assurance process known as Model Checking!
ACM has all the details in their press release.
It looks like a decent year for Advanced Scientific Computing Research at the Department of Energy’s Office of Science. Following the FY08 omnibus, in which ASCR received an almost 25 percent increase, the President has requested another 5 percent increase for FY09, for a total of $368.8 million. Here is a brief breakdown:
The high-performance computing number includes:
US high-performance computing is expected to reach a petaflop this year at Oak Ridge and Raymond Orbach, the director of the Office of Science, stated at the budget briefing presentation that the US will increase computing power by a factor of ten every two years moving forward.
Overall, the Office of Science did well in the request with $4.7 billion, an 18.8 percent increase. This keeps the Office of Science close to the ACI trajectory announced by the President in 2006. Funding levels include:
In FY08, there were $123.6 million in earmarks in a total appropriation of $4.02 billion, which the President has zeroed out in the FY09 budget request.
The President's budget request for FY 2009 is now online and we've done a quick read through to glean some numbers of interest to the computing research community. These will likely be refined over the next few days as we figure out exactly what's in there and what's not, but it's a pretty good indication of where the President's priorities are as we head into his final year.
The Networking and Information Technology Research and Development (NITRD) program
NITRD represents the sum total of the federal government's investment in information technology research across 13 federal agencies. Overall, the NITRD program would see an increase of 6 percent compared to estimated levels for FY 2008, due largely to increases in the three agencies featured in the President's American Competitiveness Initiative (ACI). IT R&D at the National Science Foundation would grow 17 percent> over FY 08 levels to $1.090 billion (putting NSF's share of NITRD at over a billion dollars for the first time). The Department of Energy's Office of Science computing research would grow 13 percent over FY 08 to $494 million. Dept of Commerce, which includes the National Institute of Standards and Technology, would grow 6 percent to $90 million.
Defense IT R&D appears to decrease 2 percent in the President's request vs FY08, but it's hard to assess that decrease without understanding exactly how many congressionally-directed projects (earmarks) were removed in the agency request. (More below.)
NASA and the National Institutes of Health also see either flat-funding or slight decreases in the request, but again, without knowing what earmarks were removed, it's hard to assess the budgets.
EPA and the National Archives and Records Administration would get what little they received in FY 08 in FY 09 ($6 million and $5 million, respectively).
Agency budgets:
NSF (pdf)
NSF research accounts would increase 16 percent (14 percent for NSF overall) over FY 08 in the President's plan, to $6.06 billion. Included in that $6 billion is "$1.1 billion for fundamental information technology research and cutting-edge supercomputing and networking resources, including: $100 million, an 110-percent increase, for an NSF-wide effort to develop radically new computational concepts and tools [this is Cyber-enabled Discovery and Innovation -- Peter]; and $30 million for a new targeted cyber-security research effort in privacy, fundamental theory and usability."
We'll have CISE numbers after NSF's budget briefing later this afternoon.
DOE (pdf)
DOE Science Programs would grow 19 percent vs FY 08 to $4.7 billion. As noted above, DOE's IT R&D would see a 13 percent increase (on top of the nearly 25 percent increase DOE's Advanced Scientific Computing Research account received in the omnibus for FY 08).
NIST (pdf)
NIST core research would increase 4 percent over FY 08, but given the heavy earmarks in the omnibus that were likely stripped from this agency request, that may actually seem like a much more substantial increase.
NASA (pdf)
NASA science would drop 4 percent to $4.4 billion.
NIH (pdf)
NIH is flat-funded in the President's request.
Defense (pdf)
This is trickiest to figure out given the how heavily the DOD budget is earmarked. The President's budget calls for an increase of just 4 percent for Defense Basic (6.1) research and a decrease of 16 percent to Defense Applied (6.2) research vs. FY 08. However, if you subtract the earmarks from the FY 08 baseline, the increase for DOD 6.1 is more like 17 percent. DOD 6.2 shorn of earmarks would also grow in FY 09 to look like a 3.5 percent *increase* over FY 08 (not a 16 percent decrease). But the devil's in the details and we'll have many more of those in the coming days.
On the whole, it looks like the President has followed through with his commitment to ACI in his final budget. Of course, he's also pledged to take some very firm stands regarding earmarks in the upcoming appropriations process (he's threatened to veto any appropriations bill sent to his desk that doesn't cut FY08 earmark levels in half). That stand virtually guarantees he won't be around when Congress finally gets around to passing approps bills. It's very unlikely Congress will want to a) give up that many earmarks and b) engage in a battle over appropriations before the election, so it's likely this won't get settled until January 09 (or later). But, as with last year, we start with some pretty healthy numbers. In fact, in terms of IT R&D, we start with the healthiest requests we've seen in many years.
More details to come.