A brief recap from Tia Newhall of tips for reading research papers.

Start keeping track of all the papers you have read in a bibtex format. You can store your notes about the paper in a field called Annote. If you have a mac, there is an awesome utility that helps you do this called BibDesk. JabRef is a useful tool written in Java. I recommend that you write your reviews in latex. Your final report will need to be in a format that is easiest to do with latex. See the department wiki page on latex to get started. Additionally, the latex source of my example review is also provided below.

Review Requirements

These requirements are also listed in the syllabus tab. Each review should be one to two pages and cover the following questions about the paper:
  • What problem did the paper address? Who is the audience?
  • Is it important/interesting? What was the context for the paper? Why should the audience care?
  • What is the approach used to solve the problem?
  • How does the paper support or otherwise justify the conclusions it reaches?
  • What are the strengths of the paper?
  • What are its weaknesses?
  • What problems are explicitly or implicitly left as future research questions?
The detail in the example review below is what you will need in the review you present in class. Your weekly review (when someone else is presenting) will follow the same format, but you only need to do ONE of the following:
  • place this paper in context of papers we haven't already read in class by actually skimming some of those papers and then citing them as appropriate in your review. (In the example, see the second paragraph for the second question).
  • provide an example to illustrate the approach
  • generate a number of future research questions that were not specified explicitly in the paper.
All three of the above activities are quite useful when reviewing a paper, but do take time.

Example Review

Here is an example review for Lamport's 1974 paper, "The parallel execution of DO loops" (pdf, tex).