-1
$\begingroup$

There are many datasets released on the internet. Authors of many of these datasets state that the datasets are strictly for academic usage and not for commercial purposes. Although some datasets are released for both academic and commercial use, many of them are restricted from commercial use.

If someone uses many of these academic datasets to train a model and then offers this trained model as a REST-API based Cloud service to earn a profit, then what is the way to find out that he or she used academic datasets to train the model? Many people might be already using academic datasets to earn a profit?

Similarly, If I collected data from many of my friends and family and published it for the academic research community and did not allow license for commercial purposes, then someone might use this data to build products and sell it commercially? How can we find that my dataset was used unethically? My friends and family might not give their data as they won't like someone to earn a profit from their data?

$\endgroup$
1
  • $\begingroup$ Please ask this question in the open-data exchange! It's more suited there! $\endgroup$ – Aymuos Nov 20 '20 at 21:30
1
$\begingroup$

If the model is a discriminative model (e.g. a classification model), it is highly unlikely that you can identify whether it was trained with some specific dataset.

If the model is generative, (e.g. a language model or a machine translation system), you may be able to try to identify if the model was trained with your data by trying to extract from it information only available in your data. This article for instance studies the feasibility of doing precisely that.

However, a different issue would be to prove that the model was indeed trained on your data.

Yet a further issue would be if such a fact is legal or not for a specific country/legal system; see these questions on the matter to better understand the problems posed by this kind of situations (I am not a lawyer, you should seek professional legal advice for this kind of matter):

$\endgroup$

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