I am interested in using data distributed under the Creative Commons Attribution-ShareAlike 4.0 (CC-By-SA) license. A subset of it can be extracted, which will fall under a Creative Commons Attribution 4.0 license (CC-By) instead.

I will be using this data for commercial, closed-source software. Only data will be used, and there is no software that will be depended on by my software.

Given that I will not be modifying or distributing the data (consumers of the software will not be able to view the data), but will only use it to automatically perform data analysis and modify another dataset that I own, what conditions would be applicable to my usage of it?

Edit: to clarify, the use of the data would result in my software being different than if it didn't use the data

  • Can you be more specific? When you say "a subset can be extracted" you mean this subset is licensed under CC-BY by the authors? Is the software that you are making and distributing going to be different because you used the data (such as an AI program)? In that case, it could be considered to be a derivative work.
    – Zimm i48
    Apr 3, 2017 at 17:36
  • The dataset in question is ConceptNet, which aggregates and links information from different sources. Part of the dataset is licensed under CC-BY. (github.com/commonsense/conceptnet5/wiki/…). I am indeed going to make AI-powered software, and make it available to customers
    – kanghj91
    Apr 4, 2017 at 0:58
  • Did you copy the correct url for the similar question? That link goes to wikipedia :D
    – kanghj91
    Apr 4, 2017 at 1:03
  • Sorry here is the link to the similar question opensource.stackexchange.com/q/5208/5858 (similar but not a duplicate)
    – Zimm i48
    Apr 4, 2017 at 10:46
  • 1
    Thank you both @Zimmi48 and @curiousdannii! So if I understand this correctly, my work will then be considered a derivative of the dataset. Under CC-BY, I will then have to provide attribution to the dataset and its authors somewhere in my software. If I use the full CC-BY-SA data, I will have to distribute my software under CC-BY-SA.
    – kanghj91
    Apr 5, 2017 at 0:59

1 Answer 1


Indeed, as was said in the comments, your software will "contain the data in a hidden way" or be better because the data was used to train it. Thus, your software will be a derivative of the data and if this data is licensed under CC-BY-SA the share-alike clause applies, whereas if you use the subset which is licensed under CC-BY, you only have an attribution requirement when making a derivative work.

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.