Licensing of AcousticBrainz datasets


Last year we released experimental feature that allows people to create datasets in AcousticBrainz. One problem is that we don’t have a license defined for them. Our data is CC0, but I think it would be nice to give dataset creators an option to ask for a credit.

We’d like to hear some suggestions and opinions on how this can work.

There was already a bit of a discussion on IRC:


Thanks for bringing this up. I agree with you that letting users ask for credit is a great idea, but it may not always be what someone wants. Perhaps we could have just 2 options - CC0 or CC-BY, and let people choose it by saying

[x] I want to be credited for this dataset.

We will have to investigate a little about what a derivative work is. If someone makes a model using a dataset licensed under CC-BY, I don’t know where the citation requirements come in.


I made this point/suggestion on IRC, repeating here for those that don’t read the IRC logs for previous discussion:

This would also depend on @alastairp’s research into what constitutes a derived work though.


2 posts were split to a new topic: Test dataset submission problem