In general, each student will be responsible for writing a review for one paper for each class meeting (although we may discuss multiple papers in a single meeting).

Reviews must be submitted by 2pm the day of the class in which the paper will be discussed (this gives me time to read them ahead of time).

Submit reviews here: https://pitt-cs3551-2020.hotcrp.com

Each review must include the following sections:

1. Summary
   • What problem does the paper address?
   • What is the main contribution of the paper? What is the key idea behind how it solves (part of) the problem it addresses? How is that different from what was done in the past?
   • Brief summary of approach, techniques used, results.
2. Strengths (at least 2 bullet points)
3. Weaknesses (at least 2 bullet points)
4. Detailed Comments
   • Explain your thoughts and give your overall opinion of the paper. Do its strengths outweigh its weaknesses? Why or why not?
   • Some things to consider:
     • Is the problem new? Is it real? Is it important?
     • What assumptions does the paper make? Are those assumptions realistic?
     • Is the solution new? Is it practical? Is it better than previous solutions (and if so, how? is it strictly better or is there some tradeoff?)? Does the solution involve new techniques, or combining existing techniques in a new way?
     • Is the system implemented? Does the evaluation convince you that the authors' conclusions are correct?
5. Discussion Points (at least 2): this wouldn't be included in a conference review, but will help guide our class meetings
   • What did you learn from the paper?
   • What questions did the paper raise for you?
   • What new ideas did the paper give you for future work? Can you think of a new/better solution? Is there a related problem that needs to be solved?
   • How does this paper connect to other papers you've read (either in this class or elsewhere)?
   • Can you use any of the ideas from the paper in your own research?

Note: this format is partially borrowed from Xin Jin's Advanced Computer Networks course at JHU.