IP protection when showing preprint work

Question

This question is in line with What is an effective way to copyright my teaching material? but about preprint work, not course material.

I produced a huge original database during a COVID break (sans any financial support) on my PhD which will not be used in the thesis, since the subject changed when I returned. It is about to be published as a book, but work on this is slow.

A prof from another department is asking to see the material to use in a class on that subject. Which I may participate but I'm not counting on this. I have no reason to distrust this other professor, but at the same time my colleagues keep reminding me of previous cases of stolen research and topics and materials that happened on that department and are widely known (from other professors, but recent, verifiable, and largely unpunished).

Just to be safe, what are some ways I can protect my work? I have under a month for preparations on this. I do not mind about my material being used as coursework, or being credited (but would be nice if it is possible to enforce somehow). I do wish to protect from competing publications.

I'm thinking about just slapping a copyright right notice at the footer and keeping a recording of the first online call where we will talk about the material/database.

Answer 1

If you are in a country that is a signatory to the Berne Convention you are pretty safe. Your work is automatically copyrighted from the date you "fix" it in some media. It doesn't need a notice, though that is a convenient warning, nor does it need to be registered. Most of the world, with few exceptions is part of this compact as detailed at the link.

The only issue would be verifying that your work actually existed in some fixed form as of a certain date. There are a lot of ways to do that, including preprints, formal publishing, etc. If you have a record with a fixed (unmodifiable) date on which the body of work existed you have a claim of priority.

The Berne Convention gives you certain, but not all, rights. In particular, others may "use" your ideas in further work (though not "derived work"). You don't "own" ideas. But you have a claim against plagiarism by "fixing" the work in some media and getting a date attached to it. You don't need to hide it to maintain your rights, though for things that can be patented, it might be wise to keep it private as patent rules are different from copyright rules.

Additionally, I think most people are honest and ethical. If you have an established date for your work, there is little risk of sharing it. And, for use in a dissertation, make sure that your advisor is up to date on what you have done and what you intend to do. Even a draft of an incomplete work is under copyright when created. Keep records of conversations to assure the history exists.

In fact, I think that being open about your intentions with the professor, using dated communications, is a good guard against IP theft. It is clear to them what and when your ideas existed. Hiding things could lead to parallel work that might step on your future publication. And, a professor has a reputation to protect and IP theft and plagiarism is a pretty serious offense.

Answer 2

This probably depends on how easy you can enforce the law in your country or in your workplace. If not so easy I would not release it to the teacher.

But in both cases, I would get a DOI identifier and call it a version. This identifier will prove in the future that you were the first to create it.

For example, the Zenodo repository provides an option, to release the database to the public with delay. So you can release it under a free license right now and provide it also to the teacher, or keep it hidden for a certain period (that's an option at Zenodo). In both cases, you are receiving a DOI, which indicates you are the original author.

Answer 3

Short answer? You probably can't, and even if you can, enforcing that right will be hard.

BUt I'm not sure that means you shouldn't share it - you need to think what the consequences of saying no are to someone more senior than you - you are basically accusing that individual of not being trustworthy, unless you can come up with a reason other than "I don't trust you not to steal it".

The long answer is complex, and probably requires a lawyer (IANAL).

But somethings to consider are that facts, measurements, and other objective truths are not subject to copyright. The compilation of a dataset can be subject to copyright, but only if judgement/creativity is involved in the compilation (and then its the compilation that is protected, not the individual facts) - any compilation that aims to be exhaustive or complete cannot be protected (as then no creativity has been expressed by the compiler).

In the UK and EU, but not in the US, databases are protected by something called "Database rights", where a database is defined as:

"a collection of independent works, data or other materials which ---

(a) are arranged in a systematic or methodical way, and

(b) are individually accessible by electronic or other means"

What exactly is protected, and what the rights are varies from country to country, but generally someone has to extract/use the entire dataset, or a substantial part, of it for an infringement to have taken place.

Something can be a serious breach of professional ethics without being illegal, of course, and getting things up on a proper repository and slaping an CC-BY-ND licence on it might help with professional (as opposed to legal) complains against someone.

Answer 4

You do not need to share work-in-progress

Given that this is work in progress, it would be perfectly reasonable for you to decline to give this information to others on the basis that it is not yet reviewed and reliable. You don't have to mention any IP issues or any concern about work being "stolen". Just let people know that this work is still in progress and in an unfinished state that is not yet reliable, so you are not comfortable sharing it with others at this stage. This can be framed entirely as an issue of academic/scientific rigour --- it is preferable to wait until work is in a reliable state before it is proliferated through academic channels.

This approach will give you time to get your work up to a state where it is publishable and you can then publish a reviewed and reliable version of the work. This may include publication of your COVID dataset and publication of any associated papers (or preprints of papers) that use that data. At that point you will be free to share it with others with the expectation that your original work is properly cited. This would include citing your COVID database if other academics use that data for their work.

Answer 5

As mentioned in another answer, protecting raw data legally can be tricky depending on jurisdiction. Depending on the goals for how the data is to be used in the course, you can provide a modified/fake version of the dataset. Options for doing this include:

1. modifying or obfuscating some values such that it's still usable for demonstrative purposes, but not suitable for publishing.
2. Adding random noise to values in the dataset such that you can still get a result that's interesting for educational purposes, but not necessarily a correct one that would be suitable for publishing
3. Generating fake samples
4. Removing some portion of the samples to make it an incomplete dataset for publishing
5. Generating a statistical result, then generating a new fake dataset from the result that can reproduce the said result.

I've done, or have encountered, all of these. I've seen obfuscation for sensitive data that can't risk being released publicly. I've generated quick fake datasets from my own unpublished real data to provide more interesting educational examples, like demonstrating how to perform an analysis. I've also generated pseudorandom data to provide interesting test datasets for software testing, documentation, and tutorials.

This approach might not be suitable if a goal of the course is to specifically explore/discuss the data in detail.

Also, make it clear that when sharing this data that it is not a valid dataset. Don't try to hide it. And in your case, discuss with the professor beforehand your concerns about sharing unpublished data and whether this would be a suitable alternative.

Answer 6

Here's what I would do

Just like you thought, putting a copyright notice everywhere on your work tells people it’s yours.

When you share your stuff with that prof, have a quick chat about how it’s cool for class but not for them to publish as their own. Maybe even write down that agreement..

Having a digital copy saved with a timestamp (like an email to yourself) can be a handy backup to prove it’s your work.

Look into Creative Commons licenses. They’re an easy way to let people know how they can (or can’t) use your work.

Keep all your emails or messages about sharing your work. Just in case you need to show proof later.