Rights and Licenses for Linked Data

Tutorial within the ESWC 2014


Venue, Date, Contact

The tutorial has a full day length and it takes place in Anissaras, Hersonissou (Greece), the 26th May 2014 (see general schedule), within the ESWC2014 Conference

Contact: tutorialrightseswc2014 --at-- delicias.dia.fi.upm.es


This tutorial is aimed at providing a basic knowledge about the existing schemas and vocabularies to represent licenses in Linked Data, and the existing formalisms able to deal with derivative data licensing and rights inferences.

9:30 - 10:30 Introduction and Use Cases (P. Casanovas)
10:30 - 11:00Coffee break
11:00 - 12:00 Deontic reasoning (N. Rotolo)
12:00 - 13:00 Session 3 (G. Governatori)
13:00 - 14:00Lunch break
14:30 - 16:00 How to license Linked Data. ODRL2.0 (V. Rodriguez)
16:00 - 16:30Coffee break
16:30 - 17:30 Applications. Exercises (S. Villata)
17:30 - 18:00Discussion


The Web of Data is assisting to a growth of interest with respect to the open challenge of representing and reasoning in an automated way over licenses and copyright. In particular, publishing and consuming Linked Data require to be aware of the terms of reuse which can be applied to the data itself. This tutorial explains what every Linked Data publisher or consumer should know about intellectual property rights and Linked Data, and it gives an advanced knowledge for acquainted users. After this tutorial, the participants will have a knowledge of which Linked Data resources are protected by intellectual property rights, how to use the well- known licenses, how to write specific rights expressions and how to reason about the licensing terms to be associated to derivative or aggregated data. In addition to each theoretical session, a short hands-on session is scheduled.


Linked Data resources have become valuable resources whose reutilization to produce new Linked Data or to power new applications is likely to grow. Moreover, the amount of available Linked Data resources has not stopped growing in quality and quantity, and Linked Data resources in the Web seem to be crying out to be exploited. However, not every use of every published resource is allowed. Under certain conditions, Linked Data resources (ontologies, datasets, etc.) can be considered intellectual property works and thus receive protection from the law. Before transforming these resources, republishing them or using them for powering applications, a public license or a private authorization allowing to do so must exist. The first session of the tutorial is dedicated to an introduction about the general problem of adopting licenses and rights in the Linked Data scenario, presenting a qualitative and quantitative analysis of the choices actually performed on the datasets of the Linked Data cloud. This introductory session includes also a discussion of the major needs of public institutions and private companies in terms of licenses and rights for Linked Data.

The fourth session of the tutorial is dedicated to the presentation of the ways to license Linked Data. The licenses offered by Creative Commons, Open Data Commons and other institutions are increasingly widespread, and the awareness on the importance of licensing a published resource is becoming almost universal. The next step in refinement might be users publishing their own personalized licenses, namely, writing rights expressions at will and not among a set of predefined licenses, for which Rights Expression Languages already exist. One of the most relevant of these rights expression languages for the Linked Data community is ODRL2.0, which has been recently published along with an ontology. Specific licenses can be created with ODRL2.0 in RDF with a great flexibility for granting rights, imposing complex conditions, and going beyond open licenses.

The second session of the tutorial is devoted to the introuof Semantic Web regulatory models, and the role of validity in the legal theory 2 . This session is dedicated to discuss the asymmetry between the Semantic Web architecture, and the architectures architectures built so far to give an account of the main components of legal theory.

The forth session of the tutorial has the aim to present formalisms and techniques to reason over licenses and rights. Once formalized, the deontic expressions (i.e., permissions, prohibitions and obligations) constituting the licenses can be logically tested against the consumer's context to determine if a certain action is allowed or not. But the scene is more complex. A common practice in Linked Data, reusing, mixing and transforming two or more resources, should take into account the license associated to each single resource. The interplay between these licenses is a matter that formal frameworks, like deontic logic, can help to decide, and if Linked Data resources are to become Lego pieces to be used once and again, it is not an issue to underestimate. Each inferred or derivative right and license must be ensured to be normative compliant with respect to the legal specification. Existing techniques to ensure normative compliance will be presented and discussed. Yet, this reasoning is not useful if it has departed from reality and lost contact with the law perspective. Formalisms like LegalRuleML can be useful and help staying close to the law domain. Legal-RuleML has been proposed to enrich the RuleML family with features specific to the formalization of norms, guidelines, policies, and legal reasoning, with a meta-model founded on RDF triples to allow Linked Data integration


Photo used under a Creative Commons CC-BY license, from oberazzi