Mary Lou Soffa, Barbara Ryder - Department of Computer Science

advertisement
Successful Fellowship
and Grant Applications
Joe Urban
Barbara Ryder
Mary Lou Soffa
Outline

NSF – Joe Urban

Grants – Barbara Ryder

Fellowships - Mary Lou Soffa
Successful Grant
Applications
Barbara G. Ryder
Rutgers University
Outline


Grants - a ‘necessary evil’
Funding agencies
– Government and Industry


Evaluation process
How to write a successful proposal?
– Best practices
– Proposal template
– NSF CAREER proposals

Sources to consult for more information
Why obtain grant
funding?

To fund your research program
Graduate students
 Equipment
 Summer salary
 Conference/workshop travel



To enhance funding provided to your
academic department and/or your research
group
To demonstrate the support of your peers
for your research
Research Funding - Govt

NSF




NIH, NASA, DOE, DHS



Main source of funding for academic CS research
Panels of peers review proposals in 1-2 day meetings in
Washington, DC
Serious funding difficulties lead to <10% success rate for
research proposals
Fit specific sub-areas of CS
Compiling for parallel, grid computing, PL techniques for
information-flow security
DARPA and DoD agencies

E.g., ONR, ASFOR
Research Funding Industry

Microsoft Research
Faculty fellowship program
http://research.microsoft.com/ur/us/nff/default.aspx
 External research & programs

http://research.microsoft.com/ur/us/fundingopps/default.aspx

Compilers, languages and runtimes
http://research.microsoft.com/ur/us/progsys/default.aspx
Research Funding Industry

IBM - University programs





Innovation awards (e.g., Eclipse Innovation Awards)
Faculty awards
Ph.D fellowships
Shared university research (SUR) awards
Open collaboration awards
http://www-304.ibm.com/jct09002c/university/scholars/awards.html
Evaluation Procedure

NSF
– Peer researchers receive ~10 proposals to
read 3 weeks before panel meeting
– Write reviews in NSF online system
– Panel meets for 1-2 days in DC, discusses
proposals and ranks them as competitive or
not-competitive

Panelists write panel summary comments for every
proposal
– NSF program managers make final choice of
proposals to fund
How to write a successful
proposal?




Be informed about major research program
solicitation announcements
Start early (2-3 months before the deadline)
Obtain and study previously successful proposals
written by colleagues
Seek advice on how to present your ideas
Give talks to colleagues/students for feedback on how
best to explain ideas
 Have others read your proposal draft to discover
unclear explanations or illogical flow of ideas

Proposal Desirables

A well-crafted proposal explains
What is the motivating problem?
 How is it being addressed?
 What is new about the proposed approach?
 How will the new approach be designed and
validated?
 What outcomes will be produced?

– Research prototypes? Research papers? Ph.D
students graduated? Algorithms embedded in
commercial products?

How will the work be accomplished?
Proposal Template (NSF)
Summary
 Tersely
motivate problem of interest and
give high-level view of your approach
 List proposed achievements of the
research
 Must state intellectual merit and
prospective broad impacts (e.g.,
educational, underrepresented groups)
Proposal Template
Introduction
 Motivate
problem(s) being studied, with
compelling argument for its importance and
potential impact
Background
 Existing
concepts necessary to explain the
research ideas
 Initial feasibility study with partial results on
first steps of the research
Proposal Template
Research description
 Provide
enough detail to show a coherent,
multi-year investigation of an interesting Q,
with multiple threads of inquiry
 Outline the proposed order of investigation
of open questions; be frank about different
possible paths that may be explored
 Tell how results will be validated (e.g.,
benchmarks, build prototype, proofs)
Proposal Template
Research plan
 Year
by year concrete milestones to be
accomplished by each person
 Emphasize the ‘fruits’ of the research
–E.g., software prototypes to be built
Related work
 Show
your familiarity with the area
 Describe how your proposed work
differs from existing approaches
NSF CAREER Program

Special program for junior faculty
– $400K over 5 years
– Can only apply 3 times to program
– Higher success rates for funding

Differences from ‘regular’ proposals





Individual vision vs collaborative research
Need to show can outline a major research investigation
-- a career path
Emphasis on researcher’s personal vision/impact
Emphasis on supporting data (e.g., supporting letters
from dept)
Need to ask successful colleagues for
example proposals
CAREER Key Points

Scope
 Neither
too small or too large
 Sequence of ideas (each a paper’s
worth) part of larger investigation
 Be able to show results already
obtained on some
Special thanks to Nasko Rountev
CSE, Ohio State University
CAREER - Key Points

Vision
 Why
is this problem important and
of potential high impact to other
researchers or computer
professionals in general?
 How is this work different from
previous work?
CAREER - Key Points

Be specific
 State
insights that will lead to novel
solutions
 Enumerate specific products that will
have impact
 Explicit plan to accomplish goals

Convince reader you will succeed
 Why
are you right to solve these
problems?
CAREER - Key Points

It’s about career advancement and
leadership in computing
Educational plans and real outreach activities
matter
 NSF keen to integrate teaching and research,
integrating research into the CS curriculum
 Impact to other scientific disciplines matter
 Already accomplished subgoals matter

Other Kinds of Grants




Travel grants for faculty
Workshop support grants
Broadening participation activities
grants
Educational (curriculum) grants

CPATH program
More Sources
NSF Grant Proposal Writing Guide
http://www.nsf.gov/publications/pub_summ.jsp?ods_key=gpg
Interesting U Wisconsin Site:
http://grants.library.wisc.edu/organizations/proposalwebsites.html
Mike Ernst’s (MIT) advice on career proposal writing:
http://pag.csail.mit.edu/~mernst/advice/careergrant.html
Your university’s grants organization often runs grant preparation
seminars and invites subscription to grant announcements
Successful Fellowship
Applications
Mary Lou Soffa
University of Virginia
CRA-W Grad Student Cohort, March 2007
Finding Fellowship Opportunities

Ask your advisor, peers
– Fellowships in your area?
– Fellowships at your university?

Search!
– Search engines
– Scholarship websites

(e.g. Fastweb.com, Scholarship.com)
Fellowship Opportunities





Government (DHS, NASA, NSF, …)
Industrial (IBM, Microsoft, …)
Foundations (Ford, Hertz, …)
University (e.g., Mellon program at Pitt)
Opportunities for Women (AT&T Labs,
Google Anita Borg, … )
When and How

Start looking early
– Deadlines: often early
(November to late February)

Prioritize
– Internship/monetary reward
– Prestige
– Amount of work required
– Your commitment/promises
Putting Together a Strong
Application

Positioning Yourself
– Make yourself someone your department will want to
“nominate”




Don’t ignore your GPA, it can matter
Publish, particularly as first-author
Finish departmental milestones (prelims/quals, comprehensives,
proposals…)
Research Statement
– Assume your readers are not familiar with your particular
research area (or sometimes even with Computer Science)!


Use clear, understandable, non-technical English
Convince the reader why your research is important
A Strong Application
(continued)
– Convey excitement / significance
– Be both forward and backward looking
How will the award help you complete your
PhD (in ways that other alternatives
wouldn’t)?
 How will the award build on what you’ve
already done (and what are your past
accomplishments)?

– Be self-contained (don’t assume the
reader will look at your webpage)
A Strong Application (continued)

Recommendation Letters
– Ask someone who knows you professionally

Curriculum Vitae
– Do not “hide” or “pad” information
– Make explicit


Authorship order
Publication quality (category types / acceptance rates)
A Strong Application (continued)

Other Advice
– Have lots of people critique / proofread your
materials

your advisor, other faculty, fellow students, friends,
even your family
– Examine prior successful applications, if
possible
Example Fellowship

Fellowship
– NSF Graduate Research Fellowship

Deadline
– November 3

Terms
–
–
–
–
Citizen of US
No more than 1 year of grad school completed
3 years
$30,000 per year, tuition, $10,500 tuition
allowance, one-time $1000 international travel
allowance
Inside the Application




General form
Letters of reference
Transcripts
Essays
– Personal history / personal statement
– Past work
– Proposed work
– Something creative…
Essays, Essays,
Everywhere…

ENTHUSIASM!
– Why do you love your work?
Helping others/other fields
 Advancing techniques/research
 Marriage of theory and practice

– Why is your work important?
– How will your work affect others?
– What is the big idea?

Do NOT get stuck in technical details
Fellowship Pros and Cons

Pros
– Very prestigious
– Often comes with summer position, travel money
– Provides a lot of freedom

Cons
– Lots of work to track down
– Lots of work to prepare
– Extremely competitive to get
– Maybe make you less serious about finding
research topic
Smaller Expenses

Finding equipment/resources;
– Ask department chair
– Ask dean’s office

Buying books
– Book scholarships
– TA discounts
Finding Travel Money


External travel funds
Conference funds
– Volunteer
– Apply for a scholarship

Ask professors, department chair, dean
– Co-authors
– Advisors
Finding Travel Money

External travel funds
– Ex:
http://women.acm.org/scholarships.html
ACM SIGs

Conferences
– May include: Registration, hotel, meals,
– Application
Advisor information
 Submissions?
 Motivation (300 words or so)

Summary

Funding is important
– Graduate student
– Faculty member
– Try to get experience in writing research
applications – grants or fellowships
Acknowledgment


Diana Litman, University of Pittsburgh
Grad Cohort, 2007
Download