Paper Summary Guidelines
Submission: by email before class
Length: ~1 page (definitely no more than 2 pages)
Suggested Content:
- Summary of the paper:
- Motivation, what is the target problem
- What was proposed and/or implemented
- What are the evaluation methodology and results
- Evaluation of the paper
- Strength/contributions
- Weakness/limitations
- Novel idea or solid implementation?
- Original thinking
- Extensions of the work
- Same/similar idea applied to other problems
Submission: by email before midnight of 03/19
Length: 2~3 pages
Suggested Content:
The main purpose of a proposal is to convince others that 1) you are working
on an important problem that is worth investigating; 2) you have a reasonable
chance of doing the work in the available time. Therefore, your proposal
should include:
- A description of the project. Include the motivation and any
necessary background that can help others to understand your proposal.
- A description of what you plan to accomplish and how you plan to get it
done.
A tentative schedule might help in communicating this.
- Any preliminary result/set-up if you have started working on this
project.
- Any backup plan in case the initial attempt fail.
Unacceptable proposals will be returned for improvement, and you will have less
time to do the work!
Important dates:
03/19: project proposal due (6%)
04/13: project progress report due (4%)
05/04: in-class presentation (10%)
05/10: final report due (20%)
Project Total: 40% of the final score
Date & Deliverables. May 4th in
class (except for Dan). A final version of your presentation slides is due
by May 5th. Note that even though the final report and associated
project material such as the source code is not due until May 10, you should
have your main results ready for in-class presentations scheduled for May 4.
Presentation Length. Each person has 25 minutes, including
questions. 20 minute presentation plus 5 minute Q&A is suggested. Time
keeping is important.
Presentation Equipment. Email your .ppt or .pdf file to
Dr. Zhong by 6pm on the day of your presentation, or bring it in a CD or
removable disk. You will only have time for minor adjustments and need to
consider the compatibility between your file and Microsoft PowerPoint for Mac.
Presentation Content. The following flow is suggested:
- Cover page. You need to have an appropriate title, author, course
information, and date.
- Introduction.
- Motivation/background to the general problem, including
previous research on the subject.
- Description of your specific research question and the key
challenges.
- Outline of the talk.
- Proposed approach/Implementation. The emphasis may vary depending
on your project topic.
- Present and justify your hypotheses if applicable.
- Key idea of your solution, system design or model description.
- Appropriate implementation details: software tools, algorithms,
issues and solutions, etc.
- Evaluation.
- Methodology: benchmarks, platforms, experiments, metrics, etc.
- Results and discussion: can you explain the results? do they
validate your hypotheses? do you learn anything surprising?
- Conclusion/Summary.
- Status: any results to be generated? which parts of the
original proposal implemented / to be implemented / cancelled? why?
- Discuss the significance of your results: any conclusions?
- Summarize your contributions in terms of tasks accomplished as well
as new skills learned. Do you achieve what you wanted to achieve?
Any special experience to share?
- Future work: possible extensions and/or applications.
- References and acknowledgements. Do not forget to acknowledge anyone who
helped in your project, including those who provided the software tool /
benchmarks.
Presentation Skill Suggestions.
- Practice, practice, practice!
- Prepare good slides with readable fonts, appropriate examples, graphs,
tables, and animations. Generally, do not write paragraphs or long
sentences and expect others to read them.
- Do not speak too fast. Give time to your audience to digest your
words.
- Make smooth transitions. Help your audience to switch between
topics.
- Stress what is important.
- Keep to the allotted time. Be prepared for dynamic adaptations.
Date & Deliverables. A PDF or WORD version of the final
project report must be submitted by May 10th.
This is a hard deadline, no extensions will be granted. If your
project involves programming effort, turn in a copy of your program with README
and necessary comments. If you perform instrumentation or obfuscation,
turn in the transformed programs as well. You do not need to submit
original benchmarks, but be sure to document them clearly. All associated
project materials should be packed into a .tar.gz or tar.zip file and submitted
by May 10.
General Advice. The final report should be a well-presented
technical writing about your project. It is the principal means by which
the project is assessed. A good final report should be detailed enough to allow
the readers to evaluate your work, yet not tedious to read. As the
presentation, you should stress what is important and highlight what you are
aimed to achieve and how you achieved them. The guidelines here are
general. Your report is highly dependent on the project topic and the work
you have done.
It is mandatory to check the spelling and grammar before submission. Contact
GMU Writing Center if assistance is
needed.
Length. About 3,500 ~ 4,500 words (excluding appendices).
This should be approximately 10 ~ 15 pages -- excluding the title page and
appendices if applicable -- for single-space format with font size no greater
than 12-point Times New Roman for the main text. Maximum number of words
is 5,500.
Plagiarism. It is essential to have
an honest and scientific attitude. Your submission must be original and
with all source quoted. Direct quotations must be indicated with quotation
marks and references. Check
here
and here for
how to recognize and avoid plagiarism. Violations of
GMU Honor Code will
result in an F.
Format. This is only a suggested format. Not all sections
are mandatory.
- Title. You should first list the title of the project, the
author, the date, the course number, and the name of the instructor.
- Abstract. A short (100 words or less) abstract that summarizes
the content and the contribution of the report.
- Introduction. Background information: what is the problem
and why you are motivated to work on it.
- Main body. The main part
that describes your proposal and reports the research effort. No
source code other than sketches of algorithms should be included here.
- Evaluation. Describe the methodology of your experiments,
report and discuss results.
- Related work. Give a brief summary of related work. It is
especially important to compare your work and alternative approaches to the
same problem proposed by others.
- Conclusions & future work. Summarize what you have achieved in
this project and discuss possible extensions/applications. It is a
good idea to refer back to your proposal at this point.
- Acknowledgements.
- References. List all previous publications that are directly
quoted or used in your report. References to these publications must
be made appropriately within the text.
- Appendices. Optional for source code, software manual, etc.
- If you failed to achieve items listed in your proposal, or
switched to the back-up plan, or changed the topic of your project, be
sure to have a special sub-section discussing the reasons and issues.
Survey Paper. If you are writing a survey paper, follow the
conventions of ACM Computing Surveys.
Check Editorial Charter
for general requirements and read papers published in that journal for samples.
The paper should provide a detailed overview of a specific area, but should NOT
be composed as a collection of paper abstracts or summaries. Your survey
should provide a comprehensive coverage of the selected area and address the
following issues:
- Introduce the importance of the area you discuss.
- Classify different approaches proposed in the area.
- Compare the classified approaches with pros/cons.
- Identify the state-of-the-art work and the latest trends in the area.
- Discuss existing limitations, open problems, and potentials.
- Extensive quoting and paraphrasing are not acceptable. All figures
and tables must be original.
- Provide novel perspectives and original insights. The originality
is an important element in evaluating the report.