Good Research Practices

@asir was asking me for resources on good research practices, and we realized there’s no central repository of information, so he suggested creating one.

I will, first dump here resources that I think communicate good research practices, and gather any suggestions. Eventually we could turn this into a informational repo, à la NLP-Progress.

Resources

Papers

Slides

Presentations

Courses

Others

2 Likes

Can we have this topic as a wiki?

The idea was to gather resources from the community, and then categorize and display the information in a more friendly format.

Do you have any suggestions on how to turn this into a wiki?

Thanks for compiling this list for us @jabalazs !
This is something that I found more targeted towards beginners: https://arxiv.org/ftp/arxiv/papers/1704/1704.01568.pdf

1 Like

Ah, my apologies for not being clear.
By wiki I simply meant this:

1 Like

Reproducing 255 papers: https://github.com/EdwardRaff/Quantifying-Independently-Reproducible-ML

2 Likes

Got it! I wasn’t familiarized with that. I just made it a wiki :smile:

1 Like

I’m writing a three-part blog series: Sane Trials of Machine Learning on a Virtual Shoestring.
It is incomplete and I sometimes revise them, so please take them with a grain of salt

1 Like

I added those three but the latter two might be too strict to beginners.

Coincidentally, I received a newsletter about paper drafting a week ago. It’s from one of the biggest paper editing agencies in Taiwan. Although the author, Dr. Steve Wallace, is an American, he wrote newsletters in Mandarin Chinese. So I took the liberty to translate its points roughly in English here.

Strategically Draft Academic Papers

One may find many “DOs” and “DON’Ts” online or via friends. Whether they will be effective are on a case-by-case basis. We think academic paper authors should develop various strategies and resources to find the best approach for themselves. Every opportunity and task deserves a different measure. Like each tool serves its own purpose, a toolbox of academic writing can’t rely on just one strategy.

Plan

Write an abstract or an outline as guidance for the drafts. This strategy works for most of the authors.

Three-Draft Cycle

Frist Draft: for the Authors

  1. See yourself and your coauthors as the earliest readers.
  2. Quick and dirty. No need to worry about writing styles for the time being.
    • Optional: sometimes be picky with the title and the introduction, because it may help clarify angles and structures, depending on the genre (conference, journal, book, etc.). The longer the paper, the more important the structure (for consistency and fluency).
  3. Pause and reflect.

Second Draft: for the Readers

  1. Delete, delete, delete.
  2. Supply and tune the introduction and the conclusions.
  3. Pause and reflect.

Third Draft: for the Publication (Story-telling, IMO)

  1. Deliver an intriguing narrative that concerns the public, not just the reviewers.
  2. Pause and reflect.
  3. Revisit the 3-draft cycle when necessary.

Conclusion

It’s neither a secret tactic nor DOs/DON’Ts tricks. It’s a method of switching among four perspectives: the planner, oneself, the target audiences, and the publication.

1 Like