Exploring compatibility between data licences
I recently wrote about work underway at the ODI to help improve the machine readability of data licensing information. We've already received some welcome feedback which has resulted in updates to the guides and the Open Data Rights Statement schema. If you've not looked at the material yet then please let us know if you have comments.
The schema, along with the publisher and re-user guides to open data licensing, provides some useful advice for anyone looking to publish or use open data. Ensuring that your rights statement is machine-readable is also a key part of achieving a higher certification for your dataset.
However there are still some areas in which we've found that people need more guidance around the implications of open data licences:
- how does an open licence affect your ability to create new, derived datasets?
- what are the implications for mixing together data and content that are covered by different open licences?
These issues are particularly relevant for people looking to build business models around open data, which often involve combining and enriching data from multiple sources. While it can be obvious how a licence might support building a product on a dataset, the implications of combining data published under different licences are not always clear.
To help address this we've put together an introductory document on "licence compatibility". This document and the supporting matrices attempt to provide:
- a summary of the key facets of the most common open data licences
- a discussion of the issues and use cases around licence compatibility
- a matrix describing which licences can be used when publishing derived datasets based upon a single source
- a matrix describing how datasets derived from multiple sources might be licensed
The document builds on previous work by the Creative Commons, extending it to include licences from the Open Data Commons and the UK public sector. The ultimate goal is to have a clearer set of guidance to support these use cases along with some tools to help guide people towards choosing the best licence for their work.
The document is still at an early stage and would benefit from review and feedback from the wider community. We would love to hear your feedback on the work to date. If you'd like to contribute then please submit issues and comments on github.