- Due Date:
- Select Paper:
Grad students, Honors contract students, others by arrangement
Reading and presenting papers is an important part of the academic life. Presenting a paper helps you by (1) forcing you to dissect the details of the paper and put each point into your own words, (2) giving you new perspectives as you discuss the paper with others, and (3) providing you a way to practice your public speaking skills.
Review the list of papers below and determine which of them interest you by doing the following:
Carefully read the title and try to determine the problem the paper attempts to solve.
Read the abstract and the conclusion sections of papers that interest you to better understand the problem, the authors’ approach, and the results from their experiments.
Sign up for the paper that interests you the most at this link by the end of the day, Thursday, Jan 31, 2019.
Presenter: Robert Wasinger*
Presenter: Gennaro De Luca
Presenter: Sohum Mendon
Presenter: Samuel Rozinski*
Presenter: William Murray
Presenter: Aditya Abhyankar*
Presenter: Anna Vuong*
Presenter: Kevin Clancy*
Presenter: Calvin Wong*
Presenter: Connor Belanger
Presenter: Eric DeJarnett
Presenter: Meagan Stephan
Michael Brand Martinez
The entries above with a star (*) next to them are the extra credit assignments. No penalty will be given if these assignments are not completed. Completed extra credit is worth up to 2% towards your final grade.
Each presenter will have 20 minutes to present the chosen paper and answer 1-2 questions. A good rule of thumb is to prepare 20 slides and spend one minute or less on each of them. Remember that when speaking in front of people, most have a tendency to take more time on each thing they want to say than they thought they would when preparing their remarks.
There is no point presenting if others cannot understand what you are talking about. Help the audience understand what this paper tries to do before showing them how the paper does it.
You do not have to present difficult technical contributions of the paper. You should digest the technical contributions and summarize in you own language and represent in a logical way.
There is never enough time to talk about everything. So you need to focus on a clear goal and message you think the paper delivers.
Use a large font in your slides. Avoid trying to cram too much into one slide and avoid reading everything on the slides. Use pictures and animations to attract the audiences.
Your slides should be neat. Use the consistent font and size. Consistency is king.
Make eye contact with the audience.
Suggested Outline: (Don’t follow strictly)
Title of the paper. Your name. Names of the authors and the universities they are from. (1 slide)
When was the paper published? Which conference or journal was the paper was published in? (1 slide)
Background of this work. Motivation and Problem Statement. Why is the problem worth investigating? How does the solution to the problem advance our knowledge? (2-3 slides)
Related Work. Prior attempts to solve the same and similar problems. (1-2 slides)
The authors’ methods to solve the problem or the question. The key idea behind their approach. The system architecture or approach workflow. (3-6 slides)
Present key results and key insights of the experiments or evaluations. Show interesting findings and takeaways. Feel free to take figures from the paper as long as they are consistent with your slides (3-6 slides)
Summary and conclusion. The impact of this paper. (1 slide)
Future directions. Your idea of how to continue working on this direction. (1 slide)
Read Section 1: Paper Presentation.
Your report should be at least 4 pages (12pt, 1.5 space, 1 in margins). You can use Word or LaTeX, but the final version must be a PDF. Your report should be neat. Use consistent font and size. Consistency is king.
The report is about explaining what this paper tries to do and how to do it in your own language.
It is also about your understanding critiques. What you think is good and bad about the proposed approach in this paper.
Suggested Outline: (Don’t follow strictly)
Title of the paper. Authors. Affiliations. Obviously, your name as well.
Introduction. Problem statement.
Background and related work.
Their approach. What do you think about this approach? What are its advantages and drawbacks?
Evaluation results. If you have a better idea to solve this problem, what kind of results would you expect from your approach?