How To Read A Paper

#How to read a paper

Don’t work your way from start to end, it’s not worthing the time.

So what should I do?

Surveying the paper

• To see if it’s even worth to read. Most of the time we search the paper based on some kind of keywords. In search engine this returns (mostly) results that contains the keywords in their title or abstract, i.e. the starting portion of a file.
  1. First thing you need to know is that you should feel free to stop reading the paper at any point.
  2. Read the title and keywords to see if it interests you at all.
  3. Read the abstract, which is the most important part of a paper.
  4. Read the conclusion. Sometimes paper were written without any outcome, sometimes the outcomes won’t help you in any way.

     Reading the article!

  5. First, take a good look at tables and figures (including captions), since it’s relatively easy to do and this would prevent you from interrupt your read.
  6. Read the introduction seriously, this is the background info and why the paper is done at all.
  7. Read the results and discussion part, they are the heart part of the paper.
  8. Stop reading, or get really deep into the details of the experiments. This is where you really learn what the authors are doing and you need to understand exactly what was done to better understand the meaning.

     Take some notes on the paper.

  9. They are not gonna serve you anything within the next week or month, but when you start to write your own paper, having some notes on these papers will be really beneficial and will certainly help you in getting your own paper in shape.
  10. Take these notes on index cards so that you can flip through them quickly and take a look at all the relevant papers.

At the end

In the next few days or weeks, I’ll post some case study using this method. I’ll be mainly reading papers that relates to:

1. Machine learning and deep learning
2. Distributed computation
3. Cloud

Thanks for reading and please feel free to leave some thought jewels: mail