# How to Read a Paper

Presented in [ACM SIGCOMM Computer Communication Review '07](https://dl.acm.org/doi/abs/10.1145/1273445.1273458). \[[Paper](http://ccr.sigcomm.org/online/files/p83-keshavA.pdf)]

Author: Srinivasan Keshav (*University of Waterloo)*

## The three-pass approach (to read a paper)

### The first pass

Give a general idea about the paper.

* **Five to ten minutes**
* Detailed steps
  1. Carefully read **the title, abstract and introduction**.
  2. Read **the section and sub-section headings**, but ignore everything else.
  3. Read **the conclusions**.
  4. Glance over **the references**, mentally ticking off the ones you’ve already read.
* Answer five questions at the end (5c)
  1. **Category**: What type of paper is this? A measurement paper? An analysis of an existing system? A description of a research prototype?
  2. **Context**: Which other papers is it related to? Which theoretical bases were used to analyze the problem?
  3. **Correctness**: Do the assumption appear to be valid?
  4. **Contributions**: What are the paper's main contributions?
  5. **Clarity**: Is the paper well written?
* Use the above information to decide whether read further.

### The second pass

Grasp the paper's content, but not its details.

* Take **up to an hour**.
* Look carefully at the **figures, diagrams and other illustrations** in the paper.
* Mark **relevant unread references** for further reading -> learn more about the background.

### The third pass

Understand the paper in depth. (Mostly a reviewer)

* **Four or five hours** for beginners, **about an hour** for an experienced reader.
* Attempt to **virtually re-implement** the paper (from the author's view).
* Identify its **strong and weak** points.

## How to do a literature survey

1. Use academic search engine to **find three to five recent papers** in this area, do **one pass** on each paper, then read their related work sections
2. Find **shared citations and repeated author names** -> obtain key papers and researchers, the top conferences