跳转至

如何阅读一篇(计算机科学)学术论文

理解

对于一篇学术论文,理解其中的技术仅仅是最基础的,但初学者往往只关注技术。对于这部分同学,我们可以问自己四个问题:

  1. Problem and Motivation:科研工作的动机是什么?
  2. Crisis in research field
  3. Weakness of existing approaches
  4. Research paradigm challenged
  5. (Claimed) contributions: (What's new in the paper)
  6. a new question asked
  7. a new understanding of the research problem
  8. a new methodology for solving problem
  9. a new algorithm
  10. a new breed of software tools or systems
  11. a new experimental method
  12. a new proof technique
  13. a new
  14. Substantiation: 作者如何证明自己的假设
  15. Conclusion: 作者的结论