4

I worked for a research project on the migration of a prototype to a low-level consumable solution. I am also co-author of a paper that describes the math used for that project (my contribution was not the math, but the tests used to evaluate the performance of the algorithm). According to the work agreement I signed, the copyright of the paper and the code belongs to university A (although it acknowledges the author's attribution).

I am also writing a thesis for an academic program at university B (paying the tuition by myself). The main researcher of university A gave me permission to use the algorithm as the basis of my individual research. I also have some ideas that can improve the overall performance.

If I base my work on the algorithm of university A (which was already published) and write a different implementation (even in a different language), could I still lose ownership of my thesis? (I am trying to avoid any infringement)

PS: the abstract of the paper of university A was accepted and exposed in a conference, but the final document has not been sent yet (as usually done in these kinds of events).

2

There are two different issues here, claims on software you write and claims on the thesis you will write.

The former depends on the jurisdiction you live in and whether the algorithm is patented or not. If you re-implement it, and can prove that you have not used any bit of the old code (which is extremely hard if you have seen the old code), then there is no copyright violation, but you might still need to pay royalties due to any patents (if your jurisdiction allows software patents). If you only have ever seen pseudo code, then there is no copyright violation as you cannot copy any code that you have not had access to. If have seen the code and you want to be sure there is no copyright conflict, ask a student who has not seen the old code to re-implement with only a natural language or very abstract description from you. Even if you are writing in a different language, if you have seen the old code, this might be constructed as copyright violation, depending on who the lawyers are and which jurisdiction you live in.

Your thesis is not covered by any copyright or patent claims on the algorithm. Copyright only covers claims on derived software. You have written a thesis which is not a piece of software. The findings in it are your own, the text are your own. Nobody can touch that. There is no patent violation either, as you are not doing anything even remotely commercial. And most jurisdictions have explicit exemptions for scientific use of patents (in others it's implicitly in the way how patents only cover commercial use).

What you should do is ask the legal department of your university to clarify this. They know how to deal with intellectual property and what is allowed and what isn't allowed in your jurisdiction. And stuff like this is their job.

  • Thanks, this is a really exciting conversation. If the code used in the University A's project was a migration of another work done by the researcher in a foreign country and the university where he developed the original code released it publicly, then I feel that the code of University A won't be a blocker for any further implementation into a third language in University B. Would a disclaimer note will be sufficient to avoid any confusions? – JFonseca Nov 28 '17 at 16:09
4

You confuse two different things: The copyright attaches to the written work (in your case the thesis). The intellectual property attaches to the content of the contribution (in your case, the algorithm).

Likely (though I'm not qualified to give legal advice), the copyright for your thesis resides with you and/or university B, whereas the intellectual property in the algorithm resides with university A.

  • 2
    Copyright is a form of intellectual property. – Dan Romik Sep 22 '17 at 9:30
  • 3
    More precisely, the specific implementation is covered by copyright. The idea of the algorithm is not copyrightable, only the actual code (so you can reimplement it and will have your own copyright). However, it might be patentable. If university A filed a patent on the algorithm, you might need a license to use your own implementation. I do not know what kind of use might be patent-relevant. – nengel Sep 22 '17 at 12:03
  • I see, WIPO says that IP covers: patents, copyright, and trademarks. And although copyright protects moral and commercial rights on the software (like @nengel said), patents exclude the commercial use to only one legal person. That is why SURF (to give an example) can only be used without a license for educational purposes. – JFonseca Sep 22 '17 at 23:50
  • @nengel put more clearly what I meant to convey. The copyright to your thesis is with university B; the patent on, or the right to patent, the algorithm is with university A. The copyright in the actual implementation of the algorithm may reside with either university, but B may infringe on A's rights if A chooses to patent the algorithm. A may have already lost their right to patent, however, if the algorithm has already been published. – Wolfgang Bangerth Sep 23 '17 at 4:07
  • Are you saying that if the code is divulged in public (by university A) before applying for a patent, it will invalidate the patentability? Clever. – JFonseca Sep 23 '17 at 23:20

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.