Hi Leigh, very interesting blog post! It's important to understand the compatibility among the existing licenses. In this context, we studied in [1] how to verify in an automated way (using a set of compatibility rules) the compatibility among a set of licensing terms associated to different licenses, expressed using the CC vocabulary. If the licenses were found compatible, then they were composed into a so called composite license (following proposals in service licenses composition [2]).
> Yes, I'm aware of L4LOD. It's essentially the same as LIMO, ccREL and, >to a lesser extent, ODRL. All of these attempt to provide terms for >describing the key facets of licenses. The benefit of ccREL is that >all of the CC licences are already described using those terms so the >machine-readable metadata exists. Concerning l4lod, I agree with you, it's actually comparable with LiMo, ccREL and ODRL. The aim of this vocabulary is to provide an alignment of existing vocabularies where licenses are mentioned, or are the focus of the vocabulary itself. We used this vocabulary to express the components (i.e. obligations, permissions and prohibitions) of the composite license we built. I understand your point about using ccREL, however if a work is not released under CC licenses, other vocabularies have to be adopted to provide the machine-readable meta-data describing the license (for instance ODRL). Finally, I have to admit that, as result of our survey on licenses used in the LOD cloud (following your previous survey), we found that almost none of the datasets expresses the licenses in a machine-readable format. Usually licenses were pointed in the VOID meta-data using the URL to the human readable license, and that's a pity… >However from a quick skim I must admit to being confused by their Real >World example in Section 3.6. While the logic might be correct in the >derivation of the combined licence, its not a great example because: >* You can't create a new derived dataset using data published under a >no-derivatives licence -- so the scenario isn't allowed >* The legal terms of the ODbL licence indicate that any derivatives >that are shared publicly must be done so under the ODbL or a >compatible licence designated by the publisher -- the derived licence >is neither Thanks for your comments! Just a quick remark about the example we used in our paper and the logic. The example does not talk about a derived dataset. The scenario we used in the paper is the following: the result of a SPARQL query against a SPARQL endpoint returns data from different datasets released under different licenses, so using our framework we provide in an automated way the licensing terms (permissions, prohibitions and obligations) associated to such query result (then called composite license) in such a way that they are normative compliant w.r.t. the licensing terms of the original single licenses. Note that our framework can be used to automatically establish whether a set of licenses are compatible or not. Further details about performances can be found in [3]. >So while there may be some value in being able to automatically create >summaries of the combined obligations/permissions of licences, I think >this is at most useful for helping understand your obligations, not >the creation of new downstream licences. Partly because it glosses >over important legal points in the terms, and partly because the >community is not best served by a proliferation of licences. >Convergence creates simplicity. I totally agree about the problem of proliferation of licenses, in fact this is not our aim. We'd like to provide, thanks to this composite license, a better understanding of the terms of reuse associated to a query result, where it's not always clear what I can actually do with the data. However, still much work to do in this direction. Cheers, Serena [1] Villata and Gandon, Licenses Compatibility and Composition in the Web of Data. 3rd International Workshop on Consuming Linked Data (COLD 2012) [2] G. R. Gangadharan, M. Weiss, V. D’Andrea, and R. Iannella. Service license composition and compatibility analysis. In ICSOC, LNCS 4749, pages 257–269, 2007. [3] Governatori G., Rotolo A., Villata S., Gandon F. One License to Compose Them All: a Deontic Logic Approach to Data Licensing on the Web of Data. 12th International Semantic Web Conference (ISWC 2013) ----- Mail original ----- > De: "Leigh Dodds" <[email protected]> > À: "Ghislain Atemezing" <[email protected]> > Cc: "public-lod community" <[email protected]>, "Serena Villata" > <[email protected]> > Envoyé: Mardi 13 Août 2013 11:12:35 > Objet: Re: Open Data Rights Statements > > Hi, > > Yes, I'm aware of L4LOD. It's essentially the same as LIMO, ccREL > and, > to a lesser extent, ODRL. All of these attempt to provide terms for > describing the key facets of licenses. The benefit of ccREL is that > all of the CC licences are already described using those terms so the > machine-readable metadata exists. > > Thanks for the pointer to the paper. > > However from a quick skim I must admit to being confused by their > Real > World example in Section 3.6. While the logic might be correct in the > derivation of the combined licence, its not a great example because: > > * You can't create a new derived dataset using data published under a > no-derivatives licence -- so the scenario isn't allowed > * The legal terms of the ODbL licence indicate that any derivatives > that are shared publicly must be done so under the ODbL or a > compatible licence designated by the publisher -- the derived licence > is neither > > So while there may be some value in being able to automatically > create > summaries of the combined obligations/permissions of licences, I > think > this is at most useful for helping understand your obligations, not > the creation of new downstream licences. Partly because it glosses > over important legal points in the terms, and partly because the > community is not best served by a proliferation of licences. > Convergence creates simplicity. > > Cheers, > > L. > > > > On Mon, Aug 12, 2013 at 5:53 PM, Ghislain Atemezing > <[email protected]> wrote: > > Hi Leigh, > > Nice work indeed! I confess I didn't go through all the guide. > > > >> This work looks at the implications of various open licences on > >> the > >> creation of derived datasets. There's a blog post with pointers > >> here: > >> > >> http://theodi.org/blog/exploring-compatibility-between-data-licences > >> > >> If anyone has any comments then please let me know. > > > > I was wondering if there were connection with the work of Serena et > > al. at > > INRIA (WIMIX team) on License composition...basically with this > > ontology > > L4LOD (Licenses for Linked Open Data) [1], and this paper [2] > > explains all > > the logic behind. > > > > > > Cheers, > > Ghislain > > > > > > > > [1] http://ns.inria.fr/l4lod/v2/l4lod_v2.html > > [2] > > http://www-sop.inria.fr/members/Serena.Villata/Resources/icail2013.pdf > > -- > > Ghislain Atemezing > > EURECOM, Multimedia Communications Department > > Campus SophiaTech > > 450, route des Chappes, 06410 Biot, France. > > e-mail: [email protected] & [email protected] > > Tel: +33 (0)4 - 9300 8178 > > Fax: +33 (0)4 - 9000 8200 > > Web: http://www.eurecom.fr/~atemezin > > > > > > -- > Leigh Dodds > Freelance Technologist > Open Data, Linked Data Geek > t: @ldodds > w: ldodds.com > e: [email protected] > >
