CISE - Advisory Committee
MEETING SUMMARY

The National Science Foundation
Directorate for Computer and Information Science and Engineering Advisory Committee (CISE AC)

April 22, 2005

National Science Foundation
Arlington, VA

Members Present:
Dr. Alfred Z. Spector, Chair, IBM Research Division, Hawthorne, NY
Professor Sarita V. Adve, Department of Computer Science, Univ. of Illinois at Urbana-Champaign, Urbana, IL
Professor Alfred V. Aho, Department of Computer Science, Columbia University, New York, NY
Professor Deborah Estrin, UCLA Computer Science, Department Los Angeles, CA
Professor Deidre Evans, Department of Computer and Information Sciences, Florida A&M Univ., Tallahassee, FL
Professor David J. Farber, Computer Science and Public Policy, Carnegie Mellon University, Pittsburgh, PA
Professor Steven Gribble, Department of Computer Science and Engineering, Univ. of Washington, Seattle, WA
Professor Barbara J. Grosz, Division of Engineering and Applied Sciences, Harvard University, Cambridge, MA
Mr. Dwight A. Gourneau, President, NAMTech, Inc., Rochester, MN
Professor Marti Hearst, SIMS, University of California-Berkeley, Berkeley, CA
Professor Elizabeth R. Jessup Department of Computer Science, University of Colorado, Boulder, CO
Professor John L. King, University of Michigan, School of Information, Ann Arbor, MI
Dr. Maria M. Klawe, School of Engineering, Princeton University, Princeton, NJ
Professor Dick Lipton, College of Computing, Georgia Institute of Technology, Atlanta, GA
Professor Joseph O’Rourke, Department of Computer Sciences, Smith College, Northampton, MA
Mr. Patrick Scaglia, Internet and Computing Platform Research Center, HP Laboratories, Palo Alto, CA
Professor Fred B. Schneider, Department of Computer Science, Cornell University, Ithaca, NY
Ms. Padmasree Warrior, Chief Technology Officer, Motorola, Inc., Schaumburg, IL (via phone)
Dr. Telle Whitney, Institute for Women and Technology, Palo Alto, CA (Ex Officio, CEOSE member)

Ex Officio Members Present:
Dr. Shenda Baker, Dept. of Chemistry, Harvey Mudd College, Claremont, CA (Ex Officio, MPS AC member)
Dr. John Wooley, Research, University of California, San Diego, CA (Ex Officio, BIO AC member)

Members Absent:
Dr. Peter Chen, Computer Science Department, Louisiana State University, Baton Rouge, LA
Professor William J. Dally, Computer Systems Laboratory, Stanford University, Stanford, CA
Dr. Linda Katehi, School of Engineering, Purdue University, West Lafayette, IN (Ex Officio, ENG AC member)
Dr. Daniel T. Ling, Microsoft Research, Redmond, WA
Professor Patricia Nava, Dept. of Electrical and Computer Engineering, Univ. of Texas at El Paso, El Paso, TX
Dr. Diana G. Oblinger, Vice President, EDUCAUSE, Boulder, CO (Ex Officio Member, EHR AC member)
Professor Rosalind W. Picard, Affective Computing Research, M.I. T. Media Lab, Cambridge, MA
Dr. David Tennenhouse, Research, Intel Corporation, Santa Clara, CA
Professor Bryant W. York, Department of Computer Science, Portland State University, Portland, OR

Senior CISE Staff Present:
Dr. Peter Freeman, Assistant Director
Dr. Deborah Crawford, Deputy Assistant Director
Michael Foster, Division Director, Computing and Communication Foundations (CCF)
Wei Zhao, Division Director, Computer and Network Systems (CNS)
Dr. Sangtae Kim, Division Director, Shared Cyberinfrastructure (SCI)
Dr. Michael Pazzani, Division Director, Information and Intelligent Systems (IIS)

The National Science Foundation’s Directorate for Computer and Information Science and Engineering Advisory Committee (CISE AC) held their spring meeting at the National Science Foundation in Arlington, Virginia on April 22, 2005.

OPENING REMARKS

Dr. Peter Freeman, Assistant Director for CISE, and Dr. Alfred Spector, Chair, called the meeting to order at 8:15 a.m. and welcomed new CISE AC member Dwight Gourneau. Dr. Spector said CISE needs the advisory committee advice in this time of budget constraints to help make sure the research balance is correct and the priority items funded.

The CISE AC approved the October 22, 2004 meeting minutes.

NSF AND CISE UPDATE

Dr. Freeman thanked the CISE AC members for their important contribution to NSF and CISE. The challenges ahead are great and the CISE field is essential both nationally and internationally to nearly everything that goes on in modern societies. Research and education is at the center. Dr. Freeman introduced the new Division Director for Computer and Network Systems (CSN), Dr. Wei Zhao.

Dr. Freeman briefly reviewed his April 21, 2005 memo with updates on CISE staff, the organization, the budget, and programmatic focus. Larry Landweber’s term as Senior Advisor to CISE will end and Dr. Freeman is looking for a replacement. The NSF budget request for FY2006 is an increase of 2.4% over FY2005. CISE’s request is 1.1% over FY2005. CISE AC meeting programmatic discussions will focus on the Theory of Computer Science, the Science of Design, declining funding rates, and the actions CISE has taken to improve them.

Dr. Freeman noted that one of the two Science and Technology Centers funded was the TRUST project in the area of cybersecurity. Fred Schneider is a senior partner on the project. The CISE AC requested data, if available, on how much funding from the Office of Integrated Activities (OIA) is for CISE-related areas. Another CISE-related STC funded was in polar research.

CISE is currently conducting a review of the CISE program areas. CISE distributed a survey to about 250 individuals (identified by the Program Officers) with a brief set of questions. The goal is to provide an evaluation of the scientific vitality of each of the current CISE program/research areas. The purpose is to provide a ten-year perspective (five back, five forward) on the scientific and educational state of computing research. A third phase of the study plans to include industry and broader review of the areas. CISE hopes to identify areas that they are currently not supporting but should be. Suggestions for additional individuals to contact were welcome. Dr. Freeman hopes to have an update at the next meeting. The CISE AC noted that bioinformatics and new applications of computer science ideas were not on the list of program areas. The CISE AC also suggested adding a few junior people to the list.

Dr. Freeman noted that Cyberinfrastructure (CI) has usually been a specific topic of discussion, but he wanted the CISE AC to focus this meeting more on the research mission of CISE. The CI issue is in a state of planning at NSF and Dr. Bement issued a process for NSF to develop a plan for what they will do in CI. At the October 2004 National Science Board (NSB) meeting, NSF was requested to provide a plan for CI activities. Planning is underway. In addition to an NSF working group, the NSF Assistant Directors (ADs) are now also officially a council on CI, chaired by Dr. Bement. Dr. Freeman said the group will incorporate input from multiple community workshops and parts of the plan may be released as early as late May for the NSB meeting.

PROGRAMMATIC UPDATE ON BROADENING PARTICIPATION

Dr. Jan Cuny, Computer and Network Systems (CNS) Program Director, provided an update on the Broadening Participation in Computing (BPC) program. She also introduced Dr. Tom Windham, the Sr. Advisor for the Science and Engineering Workforce in the NSF Office of the Director. Other CISE program officers that are working with BPC are Harriet Taylor and Caroline Wardle.

Dr. Cuny said in the first year, the BPC program is targeting traditionally underrepresented groups (women, persons with disabilities, African Americans, Native Americans, and Hispanics). She shared data with the CISE AC on projected job growth in the Information Technology fields, number of degrees for computer science majors, and numbers enrolled for high school advanced placement computer science. Trends show a decline in the number of students enrolled in computer science with a projected increase in IT jobs of 1.2M by 2012. The CISE AC talked about what might be the underlying causes for this shift and pointed out that the increase in IT jobs also includes low paying support positions. The data does not reflect skill levels. However, there will still be an increased demand for PhDs and for jobs that bridge Computer Science with other areas.

Dr. Cuny said the BPC goal is "to have all of our diverse population fully participating in computer science education and research". She noted there are variances within the groups and in the pipeline from K-12 through faculty positions. Efforts are occurring at various points within this "space" in an effort to look at broader regions rather than individual points. A workshop held in the fall of 2004 provided input for how to structure the program and identified common issues among the different groups in the CISE fields.

The BPC program has three categories: Alliances, Demonstration Projects, and Supplements. Alliances can leverage efforts, build on the success of others, have a larger impact, eliminate duplication, and can target other funding and sustainability. The first round of BPC proposals are due in June 2005 with the first grants to start in the Fall of 2005. The response to BPC has been terrific. $14M in funding is available for the first year. Dr. Cuny said they hope alliances formed (even if they are not funded) will build in future years and result in a nation-wide community with alliances on a larger scale.

Dr. Cuny asked the CISE AC for suggestions and input on the program. A fall 2005 workshop is planned with NSF and industry to look at ways programs can support each other. There are also preliminary plans for a community-wide workshop in the winter of 2006 for different communities of researchers to get together to build on what others are going.

Dr. Freeman thanked Dr. Cuny for her presentation. Dr. Grosz noted that the fall workshop was excellent and thanked Dr. Cuny and others for their role in organizing the meeting.

REPORT OF ITR COMMITTEE OF VISITORS (COV)

Maria Klawe summarized the ITR COV report held March 8-10, 2005. She briefly reviewed the COV process and noted it is an important part of the performance assessment process of NSF outcomes and goals. The ITR program that ended in FY2004 was a 5-year NSF priority area. The programmatic scope was a focus on interdisciplinary IT research and education and innovative, high-risk, high pay-off research and education. Changing foci as the program evolved from FY00 through FY04 were noted. The COV focused on awards and declines made during FY01-FY03. Dr. Klawe displayed the structure and demographics of the 35-person COV committee as well as data on ITR awards.

ITR was very successful. Projects funded seemed to be in line with what the program desired to achieve and the COV anticipates NSF will see a long-lasting benefit. NSF staff did an incredible job managing the program. Dr. Klawe highlighted the ITR COV concerns/recommendations on the ITR Process and Management:

• Additional quality mail reviews could help address the problem of assembling a strong, diverse, conflict of interest-free pool of reviewers when most of the community is submitting ITR proposals.
• To help ensure that proposers, reviewers, panels and NSF PD’s address both merit review criteria, better interpretations for "broader impact" could be offered.
• NSF should explore ways to measure (as part of the review process) proposals that are high risk, high payoff and proposals that are truly multidisciplinary.
• There should be evaluation and continuing oversight of large and medium projects.

The COV concerns/recommendations on ITR outputs and outcomes were:

• Concerns about diversity in students, leadership, and participants.
• Many "best of breed" ideas enabled ITR, new interdisciplinary areas were seeded and fueled by ITR (bioinformatics, geoinformatics, scientific computing, e-business), and the program encouraged community building by researchers and by NSF Program Directors.
• Many tools were developed and best practices are beginning to evolve. How will their impacts be evaluated and maintained after ITR? Are they now and will they be broadly accessible?
• It is critical to capture lessons learned and incorporate proven business practices to prevent future problems.

The COV found that while the ITR program had mechanisms in place to ensure meaningful and effective collaboration across disciplines, the larger proposals needed stronger management plans.

The COV recommended that in the future, large initiatives like ITR should have appropriate, assigned NSF staffing levels. NSF staff should capture lessons learned and transfer them to Program Directors who run large, complex, interdisciplinary Priority Areas. An integrated ITR web site should be established to help disseminate information and lessons learned. The ITR program has played a key role in launching interdisciplinary projects within NSF. The COV felt strongly that interdisciplinary research was important but it should be in balance with core sciences research.

Dr. Klawe recognized co-chairs Dr. Suzi Iacono and Dr. Steve Meacham who helped to organize this process.

CISE Management Response

Dr. Freeman summarized the management response to the ITR COV report, which was also included in the meeting materials. The COV was in general very positive and the ITR program has helped shaped CISE and core research in a number of areas. NSF understands that the quality of the merit review was probably substandard compared to other NSF efforts and in part, the limited oversight and insufficient reviews can be attributed to a lack of staff and the very large volumes of proposals received. Other programs within NSF have had similar concerns about addressing Criterion 2 (broadening impact) and NSF is working to address this concern. There is agreement that large awards need more post-award management. Some awards are still ongoing. Six of the eight NSF Assistant Directors participated with the COV. The review brought out the need for sustaining outcomes and finding ways to harvest all of the great software, middle-ware, applications, and programs that resulted from the ITR program. Resources are limited within NSF and this issue has not been resolved yet.

Dr. Crawford noted things that NSF is doing now to improve in the areas mentioned. There was demonstrated improvement in the quality of merit reviews from FY01 to FY03 as NSF in general focused more on addressing Criterion 2. The quality of reviewers also saw more substantive feedback on declined proposals. As proposal volume has increased, NSF continues to struggle with this. CISE has been focusing on post-award management and is working to provide Standard Operating Procedures (SOPs) for Program Officers to use. They have also been looking at ways to provide feedback to PIs and identifying management requirements for larger proposals.

Dr. Suzi Iacono said the ITR program was transformative for NSF and it is a different institution now than before ITR. Learning to work across directorates and finding ways to co-fund awards has been critical for NSF in moving toward the future. It has also been transformative for science and research communities in facing big research and education challenges in new ways.

The CISE AC strongly encouraged NSF/CISE to explore ways to harvest the data, results, and lessons learned from ITR. Dr. Freeman noted this is an important topic of discussion and worth coming back to with the CISE AC. Dr. Freeman thanked the CISE AC for their comments and encouraged them to read the reports and provide any additional comments to Suzi Iacono.

PREPARATION FOR DISCUSSION WITH DR. BEMENT

Dr. Spector asked for input on topics to discuss with Dr. Bement. The CISE AC wanted to express excitement about the success of the ITR program, the importance of Cyberinfrastructure, and note concern about the declining interest in science education on the part of students. It is important for the CISE AC to offer their support and ask for ways they can help Dr. Bement promote the NSF and to get support from Congress and other agencies.

DISCUSSION WITH NSF DIRECTOR, DR. ARDEN BEMENT, JR.

Dr. Freeman presented certificates to outgoing CISE AC members Steve Gribble, Deidre Evans, Dave Farber, Joseph O’Rourke, Sarita Adve, and Dick Lipton. Dr. Freeman also presented a gavel to Dr. Spector in recognition of his term as the CISE AC chair. Dr. Freeman then welcomed Dr. Arden Bement, Jr., NSF Director, and CISE AC members introduced themselves.

Dr. Bement provided a brief update on the NSF budget and outlook. NSF anticipates a positive impact with the realignment of the House and Senate Appropriations Committee. It does however pose a challenge, as there are now new players that have not had much interaction with NSF in the past and NSF is in the process of educating them on the Foundation. This will be another tight budget year, as expected, and there will be competition for any discretionary funding. Dr. Bement hopes that NSF can at least preserve the President’s request and perhaps do a little better.

Dr. Bement shared the general priorities and vision for NSF. It is imperative to the Nation that frontier research is a focus for NSF. In the face of new and emerging national needs, many agencies have focused more on the short-term. If there is a focus on safe science over risky science, it will leave a void/vacuum. Broadening participation is another major focus at NSF. The numbers have not changed dramatically in the past three decades. NSF initiatives in broadening participation such as LSAMP, AGEP, and CREST have put about 200,000 more women, minorities, and underserved populations in the pipeline. The research community may be on the threshold of making significant improvements in the number of students in Science, Technology, Engineering and Mathematics (STEM) PhD programs. NSF would like to see more women and minorities at the full professoriate rank. In some cases, there are highly talented minority candidates, given an opportunity to use special facilities and engage in partnerships with some of the top research universities, which can play a significant role. Dr. Bement applauded CISE for bridging these types of relationships between institutions.

There are several projects in the queue for NSF’s Major Research Equipment and Facilities Construction (MREFC) program to include observational facilities (observatories, sensor networks). These types of projects need cyberinfrastructure integrated at the beginning as part of the science model and integrated in the capital investment with outlets at major computational facilities. This is needed to convert data into scientific information and knowledge. Dr. Bement said CISE and the CISE AC would play an increasing role in the future for guidance and direction in how NSF views major facilities in the future.

Discussion:

• ITR and CI Initiatives: The CISE AC was pleased by what the ITR program achieved and how it transformed the CISE community and influenced other disciplines within NSF. They are also pleased with the comments Dr. Bement has made about cyberinfrastructure and asked what role the CISE AC can play in contributing to the efforts in support of CI. Dr. Bement said the ITR Program was a huge success and resulted in some significant projects such as GEON (Geoscience Net), the National Earthquake Simulation Network, and the International Virtual Data Grid Laboratory (IVDGL). Now more and more of the original ITR investments are moving back into the directorates. Dr. Bement anticipates the role for CISE in CI and other areas will continue to increase. There is an expectation for the NSF to keep all the science fields strong, help ensure there is a workforce for the future, keep K-12 science and engineering strong, build infrastructure, and promote innovation. Dr. Bement cautioned that all of these things cannot be done alone with only 4.5% of the Federal budget. NSF must focus on other agency support. For the CI initiative, NSF has established a council that reports to Dr. Bement and they will develop a funding model for moving forward. The CI plan needs to look to the talents and capabilities that will be needed in the future and go beyond hardware and software to include social and behavioral issues. The group is developing a comprehensive document that defines all of the potential roles the NSF could (and perhaps should) perform. They are then working to develop a planned approach for how to fill those roles within and outside NSF. The OMB, OSTP and Congress will need to review the plan. The NSB is involved in all stages in approving what NSF will do.
  
  
• Broadening Participation: CISE’s program in broadening participation shows leadership and will have an impact on the CISE field. There is agreement that this is not enough however to increase the number of underrepresented groups in both enrollment and faculty positions. How will NSF programs (like ADVANCE) do this? Dr. Bement said one of the issues is staying in the PhD post-doc experience and not getting out. PhD graduates at universities need to be supported and held accountable. Dr. Bement was disturbed by polls done last year on the perception K-12 students have of science and engineers. Only 18% responded that they personally knew a scientist or engineer. This tells us that scientists don’t talk to their community. There is a lost opportunity to engage the K-12 science and math teachers in developing programs outside the classroom.
  
  
• Proposal Success Rates: The issue of declining success rates has been a concern for several years. The CISE AC heard several proposed solutions that CISE/NSF is considering to include reducing the number of solicitations (resulting in fewer proposals received), increasing unsolicited awards, and releasing announcements versus solicitations. A challenge is to find ways that unsolicited grants can be larger and more multidisciplinary (traditionally they are single-investigator grants). Dr. Bement said NSF is seeing an increase in small cluster grants with 5-6 PIs working together. The CISE AC said the data is difficult to assess. One suggestion was to look at a fraction of credible scientists in the field and determine if they are getting funding to do their work. Some felt that new ideas come from unsolicited proposals.
  
  
• Post Doc Retention: the CISE AC encouraged NSF to look at other career stages beyond early-career to include newly tenured faculty and senior faculty. Younger people are seeing the senior people struggle with research funding and directions and this may impact their going into a research field. Dr. Bement agreed there is a need to find the right balance to build career continuity.
  
  
• Increasing High-School Participation in S&E: The CISE AC has talked about how to get more high school students interested in S&E. There is negative press in newspapers today. How can the CISE AC and NSF work together to counter some of the negative influence of the press, particular in Computer Science? Dr. Bement provided an example in the materials science fields where professional organizations have banned together and garnered resources to put on camps in the summer for young people and they are becoming oversubscribed. They have also developed a special program in curriculum development for science teachers that give them ideas for common things in the home that can be used to carry out very simple experiments (not capital expensive). The interest is there. It is a question of developing ideas to get the professional societies and community involved in working with the teachers to get it done. Many schools do not offer computer science at the middle school and high school levels. They learn the applications in math and chemistry and it is seen as taking aware from the core competency that has to be taught. Dr. Bement said it might be necessary to look beyond the classroom and offer extracurricular activities. As an example, a NIST program started in Maryland will be syndicated and offered in other schools.
  
  
• Increasing Support for Basic Science: There has been more discussion about science funding by the government and support for basic funding. There will be a hearing on Capitol Hill on May 12. What role can NSF play as an expert in understanding the process of innovation and science and the value of it and what role can the CISE AC play to achieve more focus? Dr. Bement said a number of NSF staff are engaged in NSTC committees and play a direct role in shaping policies and practices. Many are involved in that on an ongoing basis. What seems to be driving the economy is knowledge: new science, new ideas, and CI that can translate new data and information into knowledge. This is the nation’s hidden capital, which translates into economic development and progress. NSF needs to find ways to demonstrate this with the appropriate economic analysis (which SBE can do). If the CISE AC members know of any particular case examples that could help make this point it would be helpful and they should forward them to CISE or NSF.

Drs. Spector and Freeman thanked Dr. Bement for joining the CISE AC.

TRIP REPORT ON VISIT TO CHINA

Dr. William Chang, Head of the NSF Beijing Office, joined Drs. Freeman, Crawford and Zhao on a trip to China. Dr. Freeman reported on the trip to the CISE AC. The Beijing office will open in November 2005 and is the 3rd NSF international office. The mission of the trip was to understand the progress and status of China’s Information Technology research and development. The scope of the trip included government agencies, universities, institutes and centers, and industrial companies. Dr. Freeman noted his status equaled that of a "Vice Minister". He displayed a list of the government agencies visited. China also has a military mechanism that funds military research. In China, the Academy of Science is more than honorific; they operate laboratories (more than 100 with over 60K employees). The six universities visited provided were of different size and diversity. There are some remnants of centralized control of the communist regime with a parallel structure in the government at each university. The group visited Four Research Centers and Institutes: The Institute of Computing Technology; Institute of Microsystems and Information Technology; China Network Information Center; and Shanghai Super Computing Center. Dr. Freeman provided details on the facilities. Firms the group visited included Lenovo, Microsoft Research Asia, and Suzhou Industrial Park.

Dr. Freeman said the group made five presentations and attended 18 banquets. They met with 9 officers at the vice minister level, 7 university presidents or council chairmen, 5 university vice presidents, and 12 center/institute directors or deans. Observations of the group are that China’s development over the past several years has resulted in tremendous growth, heavy investment, good infrastructure, and emerging strategic views. In terms of the status of research and development in IT, it is a top national priority with more applied research. China is stronger than the US in hardware development. They appear to have a standards-based strategy, strong indigenous intellectual capability, and a modest desire for fundamental research. The observations of management and social issues is that there is a historical tendency to emulate and decision-making is traditionally top-down. These may become limiting factors for sustained growth

Dr. Scaglia said Hewlett Packard has observed that the universities in China drive standards, which is different from the US. Universities are probably still under control of the Chinese government. Dr. Freeman said his understanding is that many of the top government leaders come from a technical university-based background. Universities and Science Academies seem to have a strong influence. Dr. Zhao said that the "safe" studies were in science during the revolution and they are now seeing these people in leadership roles. Dr. Freeman noted that China has more people in K-12 than the US has in total citizens. Just 1% of those students going into S&E fields would be a significant number.

In summarizing the observations, Dr. Freeman said China has made significant progress in the last 15 years with high potential to continue growth. There are significant management and social hurdles to overcome to achieve consistent world-class status. Recommendations for the CISE AC were to:

• Encourage contacts between Chinese and American scientists and engineers to improve a mutual understanding and increase awareness of the types of research activities going on in China.
• Engage in collaborations that promote academic freedom
• Consider new programmatic responses from NSF

Dr. Freeman added that all of the leaders at the top levels expressed nothing but great desire for further cooperation at the level he mentioned.

NEXT CISE AC MEETING

Dr. Freeman asked for CISE AC input on the two-day fall meeting October 20-21, 2005 at the Computer Museum in Mountain View, California. The meeting will include a guided tour of the museum and have representatives from large ITR and STC projects on the west coast. He hopes the meeting will provide an opportunity to look in more detail at some of the activities that CISE is supporting. Information will be distributed ahead of time with an informal dinner planned the evening before and a more formal dinner on October 20th.

DISCUSSION OF THE COMPUTER SCIENCE THEORY PROGRAM - PAST, PRESENT AND FUTURE STATUS

Michael Foster, Division Director, Computing and Communication Foundations (CCF), provided an update on a core CISE area and the impact of NSF/CISE budgets and the continuing CISE reorganizations on Theoretical Computer Science. There is evidence of community dissatisfaction and advice from the CISE AC is requested. CS Theory can help answer hard foundational questions, provide linkages between disparate fields and serve as an incubator for other areas. The Theory of Computing program (TOC) supports a lot of science along the risk spectrum (from NSF’s point of view). In additional there is theoretical work in CS supported by other NSF programs and the ITR program. TOC has funded many CAREER grants. The current TOC portfolio includes core theory in computational complexity; fundamental algorithms combinatorial, approximation, parallel, online, graph algorithms) and application-specific theory that includes molecular biology, networks, and linguistics. NSF has funded most of the researchers published in CS theory.

Each cluster within the CISE CCF division has a program with teams of Program Officers and staff working closely with the community. CCF has three clusters: 1) Theoretical Foundations. 2) Foundations of Computing Processes and Artifacts, and 3) Emerging Models and Technologies for Computation. Additional data was provided on the number of awards made each year in the cluster areas and through CAREER. By having Program Directors working in clusters, CISE can get the benefit of several program officers working together and looking at a broad set of proposals. The TOC program strategy is to maintain strong performers in complexity, and to narrow system-theory gaps in algorithms, and applied theory.

Questions posed to the CISE AC:

• At the Division level, should they modify the speed of clustering? Should they recluster?
• TOC Program - Should they maintain strong performers for hard foundational questions? Should they continue strategy of diverse awards? Should they encourage "impure" theory to link fields? Should they encourage co-funding to incubate?

CISE is currently looking for a program director in Computer Theory. Dr. Freeman added a general comment that other programs can co-fund projects in the core.

CISE AC Discussion:

• Though ITR resulted in moving forward joint funding resulting in many successes, some felt there should be more focus on open-ended funding in the next 3-4 years in all core areas of computer science.
• There is a need for stronger theoretical and intellectual work in all of the CISE fields, and advised caution in stating that CS Theory is THE intellectual core of the field, though it is highly valued.
• More data is needed on who is applying for grants in CS Theory. The numbers presented do not give a good sense of what the crisis is. Data on the level of funding in various areas and types of universities that are applying is needed. The CISE AC needs better/different data in order to provide advice and requested leadership from NSF in identifying some type of end-to-end measure of funding in the field.
• The loss of DARPA funding for universities has contributed to the funding crisis.
• It may be helpful to have a discussion at a future CISE AC meeting on the frontier objectives for CS Theory with data available to support the discussion.
• The group underscored the importance of core research. The CS Theory community is asking for more funding, but other areas may need more funding too. What is the basis for making the decision of who gets funded? The current mode of doing research in CS Theory is unconstrained individual grants. To get funding you need larger, grouped research and for this group, this is not the way to get things done. Theorists may argue that there is a unique style of doing research for theory and more funding for single investor research is needed.
• The CISE AC requested a list of where theoreticians in the top-30 research universities get their funding. It was suggested to also look at student enrollment, not just faculty. Who is not getting funding and what problems is that causing?
• The CISE AC offered to work with CISE offline. The issue extends beyond CS Theory because it is really about working on the frontier and how to do that.

Dr. Freeman summarized that CISE understands the need for data and they will take a renewed look at what they are doing in the core. The core of computer science cannot be only one area. Dr. Spector said the discussion on CS Theory and frontier research is likely to be ongoing.

DISCUSSION OF ENCOURAGING HIGHLY INNOVATIVE RESEARCH AND REVIEW PROCESS IN GENERAL

Dr. Crawford presented data on the number of proposals and reviews per year from FY200 to FY2004. There is an increasing reliance on panel reviews. In looking at the quality of the reviews in recent years, there are less substantive comments received. There also has not been an increase in NSF staff to support the increase in the number of proposals received by NSF. Dr. Crawford proposed several options for improving merit review quality to include: increase the use of mail reviews; reduce the number of proposals assigned per panelist; and educate the computing community (to participate in the merit review process, provide substantive feedback on reviews, and provide objective ratings).

Observations for supporting high-risk, high pay-off projects indicate that there is a culture of destructive criticism; the panel review process can create conservative group-think; independent mail review has no baseline for comparison; and panel/mail combination create workload demands. Options to increase the number of high-risk, high pay-off projects include no rankings for panel review; no budget restrictions (however it would have a consequence in the proposal success rates); conduct larger award competitions within Divisions, CISE-wide; and increase the number of Small Grants for Exploratory Research (SGERs).

The CISE AC discussed the options proposed. They cautioned NSF/CISE that there is not a "one size fits all" solution. Identifying incentives in the research community are important. Dr. Freeman said that NSF has done quite a bit of analysis on incentives. The behavior in the research community is causing some of the problems and NSF has looked at ways to manage the behavior. The culture at research universities needs to be addressed. One member noted the complexity of the issues. Is there enough funding and adjustments are made in who gets funded or is there a need to identify how much is needed and what isn’t being done? Dr. Freeman noted this is an example of the kinds of science policy discussions that go on in NSF endlessly.

Dr. Grosz mentioned there is literature that reviews conducted under time pressure result in lower success rates for minorities and women. Consideration should be given to this when talking about broadening participation. Also in some areas of research, if there are across-the-board budget cuts, there is a gender imbalance.

The CISE AC members were provided data on the CISE investments in early faculty award program CAREER grants. The CISE AC recommended that data be categorized to identify research fields within CISE as well to better identify trends. Data seemed to indicate that faculty are receiving CAREER awards much closer to tenure than early in their careers. The CISE AC requested additional data on the average number of panels that reviewers participate in and discipline-specific data on CAREER awards, if possible.

REPORT ON COMPUTING RESEARCH ASSOCIATION (CRA) SUMMIT MEETING

Dr. Spector said the CRA held a computing summit meeting February 28, 2005 attended by influential people in CISE fields. The summit convened out of a concern for funding in the field, a concern for US competitiveness, and the concern that people are not entering the field and PhD programs. The attendees talked about root causes and identified a few categories of actions:

• The marketing message needs to change the perception of computing
• Specific strategies to attract K-12 students to computing are needed
• Mechanisms are needed to address Congress and other public policy makers with respect to funding and public policy issues.
• CISE fields need to reach out to other disciplines and come up with an integrated message to "sell" science and engineering. They are not isolated fields.

A result of the summit was the funding of two task forces to address the image/public relations and policy. CRA, the Computer Society, and ACM have to get involved with this and communicate the issues through the press and editorials. The complete summary is on the CRA website at http://www.cra.org/Activities/summit/home.html.

REPORT ON THE WORLD ECONOMIC FORUM, DAVOS, SWITZERLAND

Dr. Debra Estrin attended the World Economic Forum held in Davos, Switzerland, January 29-30, 2005. The forum brings together young world leaders. Dr. Estrin said her impression was that Information Technology does not play the same role that it used to play globally and was disappointed that the “big shots” in IT that were at the summit meeting appeared to have a marginal role and technology was not integrated into discussions. She said it was a learning experience. Topics presented included telecommunications, robotics, and biotechnology.

WRAP UP

The new chair, Dr. Alfred Aho, asked the CISE AC to submit topics for the October 2005 meeting to him via email. Topics mentioned included broadening diversity; improved publicity in educating society about the CISE field; and reexamining the size of the CISE research portfolio.

Dr. Freeman thanked Dr. Alfred Spector for his service as the CISE AC chair.

Dr. Freeman thanked the CISE AC members for their contributions. Dr. Spector urged the committee and the Foundation to address the issue of integration of the CISE field and its impact on other fields to include social science, economics and other disciplines. He urged the group to come up with some ideas that could capture the imagination of the country and result in new hybridization for all of the sciences.

With no further discussion, the meeting adjourned at 3:15 p.m

CISE AC April 22, 2005 Action Items

• The CISE AC requested data, if available, on how much funding from the Office of Integrated Activities (OIA) is for CISE-related areas.
• Provide an update on the CISE Program Area review/survey at the next CISE AC meeting. It was noted that bioinformatics and new applications of computer science ideas were not on the list of program areas. The CISE AC also suggested adding a few junior people to the list.
• Additional data was requested on the average number of panels that reviewers participate in and discipline specific data on CAREER awards, if possible.
• It may be helpful to have a discussion at a future CISE AC meeting on the frontier objectives for CS Theory with data available to support the discussion. Data is needed on the level of funding in various areas and types of universities that are applying. The CISE AC needs better/different data in order to provide advice and requested leadership from NSF in identifying some type of end-to-end measure of funding in the field.
• The CISE AC requested a list of where theoreticians in the top-30 research universities get their funding. It was suggested to also look at student enrollment, not just faculty. Who is not getting funding and what problems is that causing?