6

I am submitting a paper to a computer science conference. In this paper I compare experimentally a method A (mine) with a method B (from other people).

I found the code for method B on git-hub by an anonymous author, and seems ok, with very few editing. Now, I have to upload the code I used for the experiments for review, and I don't know what to do with the code for method B:

  • If I submit it as is, the reviewer may realize that I have taken it from an "untrusted" source, or being the author of the code, who knows.
  • On the other hand it seems ridiculous to me to change the name of the functions and of the variables just to make the code "more mine".

Is there a prassi in this case?

  • 2
    What license is the code released under on github? GPL? BSD? If there is no license stated, usual copyright laws apply, which would prohibit modification and republication, in which case you probably can't upload their code. – AJK Feb 21 '17 at 22:58
  • I would upload the code for method A and not upload the code for method B. In the end, you are going to talk about your method and how it compares to an existing method. The existing method is not for them to review, it is yours that matters. – skymningen Feb 22 '17 at 13:19
7

I don't think the fundamental issue here is how to submit the code, but rather how to give credit in the paper.

How did you describe the code in your paper? If your experiments use other people's code, then you need to indicate this explicitly and give them credit. Even if the author is anonymous, it's important to indicate that you are not taking credit for it yourself, as well as to indicate where readers can obtain the code.

If you have already dealt with this in the paper, then you have no problem when it comes time to upload the code for review, since the reviewers will already know exactly what to expect.

If you haven't dealt with this in the paper, then you have a problem and will need to revise the paper ASAP.

As for whether it's from an untrusted source, you'll need to be upfront about this. If you can justify why this is a reasonable choice (for example, you've checked the code reasonably carefully), then you should do so. If you can't justify it in a way that reviewers will accept, then you have a problem, but you shouldn't try to cover it up.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.