Some colleagues and I have clicked several points from articles of the literature to retrieve the data using a tool like WebPlotDigitizer. I usually try to contact the authors to retrieve the original data. But sometimes it comes from old publications and authors cannot be reached any more. We wanted to make those data available to the community to extend the validation test cases that everyone could run. Unfortunately, I was not able to figure out the copyright status of such data. Imperial College has some information here, but not specifically on this point. It looks to me this is a grey area. It is a different case than this question, as I do not particularly want to publish the data in a journal. Does anyone have a clue on the copyright status of those clicked data? Am I allowed to put them in a CSV format on a public repository?
-
3I strongly recommend contacting a lawyer. Copyright law is complicated and is different in different countries.– Peter FlomCommented Apr 25 at 12:43
-
1Can you clarify what you mean by "clicked data"? Perhaps there's a more conventional phrasing for what you mean.– Daniel R. CollinsCommented Apr 25 at 13:57
-
1Even if the data isn't copyrighted, I wonder if EU database rights might apply.– AnyonCommented Apr 25 at 14:48
-
3@DanielR.Collins WebPlotDigitizer is a piece of software for reconstructing a table/spreadsheet of data from a graphical representation of those data - a key part of the workflow for using it is clicking on points on the graph, hence "clicked data".– Daniel HattonCommented Apr 25 at 14:54
-
Thanks for the comment! Thanks Daniel Hatton for the details about the tools. Yes Peter Flom, I will contact a lawyer in my company, but I was curious if anyone already did this. And thanks Anyon for this link! That seems quite appropriate. I will dig and come back to close this if I get an answer (which might be country dependent but could help other people).– LalyluleloCommented Apr 28 at 7:43
1 Answer
The comments already mention talking to a lawyer and that's probably the safest answer. I'll throw out my thoughts though, in case it is helpful for others.
In the US data is not copyrightable itself. So you should be free to extract and distribute the data, assuming you obtained it through legal means and did not specifically sign an NDA or other agreement that prohibits extraction/distribution. This is doubly true since you're not distributing their original dataset but a reproduction based on their published work. This is different in the EU, where database rights are a thing - though I do not think they are applicable at all to the US. If you are in the EU or other countries that do have database rights, you will likely need to obtain permission.
Seperate from the strictly the legal aspects, in biomedical sciences at least, there is a well defined methodology that extracts data from published papers and re-analyzes it - a meta-analysis. This is an expected use of published material with decades of precedent. I would say that the act of extracting and using previously published data in academia is fairly common. Meta-analyses are not a US-specific phenomenon, so I have to assume that they are broadly permissible in spite of "database rights" laws in other countries - though I don't recommend using this as a justification for ignoring those laws or skipping on legal advice if you happen to be outside the US.
-
Thanks for your answer. You provide a good point with meta-analysis and the fact that it's not the original dataset. I'll try to reach lawyers, and I'll update this question if I ever obtain an answer. Commented Jun 16 at 8:23