Whenever re-using open data from third parties, for further consuming, integration or publication purposes, and whenever such data are released under different licences, a licence comparison activity is necessary in order to check and access the legal compatibility of the actions permitted, forbidden or required by their licenses. In the specific domain of data integration, for example, all the licences attached to the datasets that should be integrated must at least permit to adapt (e.g., to transform, link, mash-up, combine) or derive from the original works a new work (i.e. the integrated dataset). Otherwise, any integration would be impossible.
Another step to take is to choose a suitable license that should be attached to the newly created dataset, once the integration is done and the new work should be released and published. Some open data licenses are aware of the necessity to use and manipulate Information coming from several Information Providers and foresee a mechanism to list multiple attributions, by considering the possibility to include a URI or hyperlink to a resource that contains all the required attribution statements (e.g. a list of multiple licenses, provided they are “compatible”, as it is intended in the following paragraphs).
COMSODE methodology will deal with these and other aspects of comparing different licences to check their compatibility that builds upon the studies of Villata & Gandon (2012) and Governatori et al. (2013). Compatibility rules have been defined over a basic model of licenses that relies on the structure of the Creative Common licenses RDF schema.. A series of step by step rules are given to compare elements of licences and a tool called “Licentia” is used as the computational device for the licenses compatibility assessment based on the above rationale.
Licentia allows to:
- find a custom license for one work, provided the permissions, obligations and prohibitions that should be associated with the use and reuse of the data. Licentia returns the list of licenses compatible with the selected conditions.
- check if one license allows for a certain use or reuse of the data. By choosing one license in the tool repository and a set of conditions. Licentia tells whether the license is compatible with the selected conditions.
- download a license in RDF (Turtle syntax) to attach it to data. The license can be also visualized as a graph-based representation.
For example, by describing one license into the system through its permissions, requirements and prohibition, a list of compatible licenses is returned by the tool. The following example is based on the search and retrieval of a list of compatible licenses for the The Creative Common License[i]:
The Figure above is an excerpts of the list of full compatible licenses as returned by Licentia (the description of terms for the Creative Common License against which the compatibility has been computed is reported on the right of the Figure).
The above figure shows an excerpt of the list of compatible licences with missing (unspecified) elements with respect to the Creative Common License.
The above figure shows a detailed list of unspecified elements of two licenses with respect to the Creative Common License.
The green elements represent fully compatible licenses, whereas yellow elements represent still compatible licenses, even if some elements of particular license needs to be checked against unspecified elements of another license. A set of rules fully reported in COMSODE methodology, states the compatibilities among elements vs. their unspecified counterpart. In the example, both the GNU General public Licence v2.0 and the Boost Software Licence 1.0 are compatible with the Creative Common License above.
A practical example
During COMSODE platform publication process, most Italian datasets have been licensed with CC-BY 3.0 Italy. By looking at the License visualizer Tool of Licencia we may discover all the terms of the license that we can compare with the license of a dataset of another country that we want to integrate with the Italian one, for example the Czech dataset with licence Open Data Commons Public Domain Dedication and License.
The two figures below show the License Visualizer Tool of an except of both the two licenses. To check if they are compatible and if we can then proceed with the integration of datasets we may use the Check compatibility Tool of Licentia.
In the figure above an excerpt of the CC-BY 3.0 Italy License – part Obligations
In the figure above the ODN Public DDL License
By running the License Compatibility Checker tool of Licentia we discover that the two Licenses are compatible. By checking the compatibility it turns out that the Attribution term is compatible with the Unspecified one.
As witnessed from the above examples, a more sophisticated theory is necessary to deal with the composition of licenses, a matter that is still debated in the legal domain, as the combination of licences is not always licit and not always create a meaningful hybrid. For these reasons, composition of licenses will be included in the scope of the COMSODE project only in the minimal required range.
- Governatori, G., Rotolo, A., Villata, S., & Gandon, F. (2013). One License to Compose Them All. In The Semantic Web–ISWC 2013 (pp. 151-166). Springer
- Villata, S., & Gandon, F. (2012). Licenses Compatibility and Composition in the Web of Data. In COLD.
- [i] https://creativecommons.org/licenses/by/4.0
Angela Locoro earned her Master Degree in Modern Literature from the Faculty of Humanities and Bachelor Degree in Computer Science from the Faculty of Science of the University of Genova, where she also attended the PhD Course in Electronic and Computer Engineering, Robotics and Telecommunications. In 2014 she joint University of Milano Bicocca working on COMSODE project.