research proposals

24
Research Proposals Computer Science Research Practicum Fall 2012 Andrew Rosenberg

Upload: kamali

Post on 22-Feb-2016

44 views

Category:

Documents


0 download

DESCRIPTION

Research Proposals. Computer Science Research Practicum Fall 2012 Andrew Rosenberg. Proposals. Who is the audience? What is the purpose? What are some standard and useful structures?. Who is the audience for a research proposal?. F amily and friends A boss Other administrators - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Research Proposals

Research Proposals

Computer Science Research Practicum

Fall 2012Andrew Rosenberg

Page 2: Research Proposals

2

Proposals• Who is the audience?• What is the purpose?• What are some standard and useful

structures?

Page 3: Research Proposals

3

Who is the audience for a research proposal?

• Family and friends• A boss• Other administrators• A potential client• A potential funding agency• A professor/advisor• A colleague

Page 4: Research Proposals

4

What is a proposal?

• A description of what you plan to do.

• The length, and required detail vary by audience.

Page 5: Research Proposals

5

What is the purpose of a proposal?

1. Communicate your plans.2. Advocate for your effort3. Compete against other proposals for

limited resources4. Gather feedback5. Clarify your thoughts.

Page 6: Research Proposals

6

What should a proposal do?

1. Motivate the problem

2. Motivate your solution

Page 7: Research Proposals

7

Common structure of a research proposal

1. Abstract2. Introduction3. Related Work4. Pilot Studies and Previous Work5. Proposed Effort6. Previous Successes7. Conclusion8. *References and Appendices

Page 8: Research Proposals

8

Abstract• An “abstract” is an high-level

summary of the proposal. • Maximum 1 page.• Minimum 1 paragraph.• It lacks many details.• Summarizes the major themes.

– Motivate the problem.– Motivate your answer.

Page 9: Research Proposals

9

Example AbstractThe Research Practicum class has traditionally left students unprepared for PhD programs. There is a need to train graduating Master’s students in fundamental research skills including but not limited to: design and implementation of experiments, presentation and writing skills. The proposed 12-week course will address this gap through hands on lectures. Students who take this course will have a significant advantage and important skill set should they continue in their academic careers into research in computer science.

Page 10: Research Proposals

10

Introduction• The introduction lays out the problem.

– What is your research addressing?– Why is it a problem?– What are you going to do? (very brief)– What will solving this problem accomplish?

• Convince your reader that there is a need for this research, or other effort.

• Be honest, but you don’t need to point out possible risk at this point.

Page 11: Research Proposals

11

Example IntroductionThe Research Practicum class has traditionally left students unprepared for PhD study. While it has been a capstone course for many years, our graduating students who have continued on to PhD study have reported that they are untrained in basic research skills. They don’t have the hands on experience in the mechanics of executing, writing and presenting research. One student reports, “I didn’t know where to begin, everyone else knew how to give a presentation, and I really never had”.

The CS department has a capstone course on the curriculum, yet it is not sufficiently training students in the skills that they will need moving forward. The proposed course will address this skills gap by providing hands on experience in three areas.

1) experimental design and execution2) writing of proposals and technical reports3) preparation and delivery of oral presentations

With training in these core areas, our students will be well prepared for successful academic careers to an unprecedented degree.

Page 12: Research Proposals

12

Related Work• What have other people done on this

topic?

• How does it relate to your proposed work?

• Where is it still lacking? – This can help motivate your solution.

Page 13: Research Proposals

13

Example Related WorkThe Research Practicum class has been taught many times before. In these courses, experimental design and execution has been highlighted. Students got hands on experience in the practice of research in computer science. For example, in Fall 2010, Prof. X had all of the enrolled students work together to improve speech recognition for non-native speakers. This experience gave students a unique experience of the realities of doing cutting edge research. However, students did not need to write a report on their efforts.

At XYZ University, there is coursework similar to the one I am proposing to teach. At this university, each of these skills are taught in 4-week courses by the Engineering school. Students do not get credit for these courses, and participation is optional. This limits the enrollment and impact of these courses.

Page 14: Research Proposals

14

Pilot Studies and Previous Work• Have you done any work previously that

relates to the proposed effort?

• What are the results of this research?

• Small applications serve as a proof-of-concept.

• More comprehensive work can serve as a testimony of your ability to execute the proposed work.

Page 15: Research Proposals

15

Example Previous WorkProf. Rosenberg has taught 11 courses in the CS department at Queens College and the CUNY Graduate Center. In each of the Machine Learning and Spoken Language Processing courses, oral presentations and research projects have been part of the syllabus.

Prof. Rosenberg has informally advised dozens of students on presentation and paper writing skills. He has written over 30 peer-reviewed papers. He has given oral and poster presentations at international world-class conferences including: NAACL, EMNLP, and Interspeech. He has given invited talks at ETS, AT&T, USC.

Page 16: Research Proposals

16

Proposed Effort• Get into the details of what you

propose to do.• Be as explicit as possible, despite the

fact that the work isn’t done yet.• A detailed plan will give the reader

confidence that you know what you are doing.

Page 17: Research Proposals

17

Example PlanIn this section, we lay outline the coursework for the Fall 2012 Research Practicum course. A week-by-week syllabus will serve as a backbone for this description highlighting the skills that will be covered, activities and any homework assignments.

Week 1 – Defining Computer Science and Research

Week 2 – Numerical Skills

Week 3 – Evaluation and Machine Learning

Week 4 – Research Proposals

In addition to this syllabus, guest lecturers will be invited to address the course. This will provide insight into the research process from people who are actively engaged in it.s

Page 18: Research Proposals

18

Previous Successes• This allows you to taut your own

accomplishments. • This motivates you as the person

who should do this work.• What have you done elsewhere that

is relevant to the success of this project. – This is different from prior work. Here

the work doesn’t need to be directly relevant.

Page 19: Research Proposals

19

Example Previous successesProf. Rosenberg has been an active member of the Queens College Computer Science Department faculty since joining in Fall 2009. He is committed to fostering research in computer science at Queens College and CUNY in general. Prof. Rosenberg’s Speech Lab has successfully secured funding from NSF, DARPA, Air Force, and IARPA. In addition to establishing this research lab, Prof. Rosenberg has mentored 5 undergraduate and master’s students through independent studies. He has also sat on the Graduate Academic committee at Queens College and the Executive Board of the CS program at the Graduate Center.

Page 20: Research Proposals

20

Conclusion• Tie everything together by

summarizing the content.• A conclusion is often like the

introduction or abstract. • Avoid bringing up new ideas in the

conclusion.• Avoid “In conclusion,”.

Page 21: Research Proposals

21

Example Conclusion

Page 22: Research Proposals

22

References and Appendices• Any papers or work that you have

referenced in the writing should be included in a References section.– There are many valid and useful reference

formats, but be consistent.– sometimes the audience will expect a

particular format.• If there are details which are too

cumbersome or tangentially important, these can be put in an appendix.

Page 23: Research Proposals

23

Example References• A. Rosenberg, “Why I teach”, Journal

of non-existent stuff, Vol 1, No. 1, 112-112

• A. Rosenberg, “Symbolic and Direct Modeling of Prosody for Classifying Speaking Style and Nativeness” Interspeech 2012

• etc.

Page 24: Research Proposals

24

Example AppendixIn this appendix, we include student feedback from Prof. Rosenberg’s previous courses.

“…not that bad…”“I’m not sure that I learned anything, but I’m pretty sure I didn’t get any dumber by taking Algorithms with Prof. Rosenberg.”“Snoozefest”