Ten Questions for Reading a Paper

Recently, I have been trying to read papers. Because my own foundation is still insufficient, I often try to skip the difficult parts and fail to absorb the essence well.

I recently saw Harry Shum's "Ten Questions for Reading a Paper" and found it very helpful, so I excerpt it here:

Ten Questions for Reading a Paper

• Q1 What problem is the paper trying to solve?
• Q2 Is this a new problem?
• Q3 What scientific hypothesis is this paper trying to verify?
• Q4 What related work exists? How should it be categorized? Which researchers in this field are worth paying attention to?
• Q5 What is the key to the solution proposed in the paper?
• Q6 How are the experiments in the paper designed?
• Q7 What dataset is used for quantitative evaluation? Is the code open source?
• Q8 Do the experiments and results in the paper support the scientific hypothesis being verified?
• Q9 What exactly is the contribution of this paper?
• Q10 What comes next? What work can be explored further?

A Few Extra Notes

• The boundaries between questions need to be distinguished correctly. The answer should hit the key points without circling around the same words.

• Try to use the paper's own expressions and help readers without prior knowledge understand the topic.

• The ten answers should clearly express the main work and contribution of the paper.