Most policy-based access control frameworks explicitly model whether execution of certain actions (read, write, etc.) on certain assets should be permitted or denied and usually assume that such actions are disjoint from each other, i.e. there does not exist any explicit or implicit dependency between actions of the domain. This in turn means, that conflicts among rules or policies can only occur if those contradictory rules or policies constrain the same action. In the present paper - motivated by the example of ODRL 2.1 as policy expression language - we follow a different approach and shed light on possible dependencies among actions of access control policies. We propose an interpretation of the formal semantics of general ODRL policy expressions and motivate rule-based reasoning over such policy expressions taking both explicit and implicit dependencies among actions into account. Our main contributions are (i) an exploration of different kinds of ambiguities that might emerge based on explicit or implicit dependencies among actions, and (ii) a formal interpretation of the semantics of general ODRL policies based on a defined abstract syntax for ODRL which shall eventually enable to perform rule-based reasoning over a set of such policies.
The integration of heterogeneous data is a reoccurring problem in different technical spaces. With the rise of model-driven engineering (MDE), much effort has been spent in developing dedicated transformation languages and accompanying engines to transform, compare, and synchronize heterogeneous models. At the same time, ontologies have been proposed in the Semantic Web area as the main mean to describe the intention as well as the extension of a domain. While dedicated languages for querying and reasoning with ontologies have been intensively studied, specific support for integration concerns leading to executable transformations is rare compared to MDE.
Based on previous studies which relate metamodels and models to ontologies, we discuss in this paper synergies between transformation languages of MDE, in particular Triple Graph Grammars (TGGs), and Semantic Web technologies, specifically OWL/SPARQL. First, we show how TGGs are employed to define correspondences between ontologies and how these correspondences are operationalized in SPARQL. Second, we show how reasoning support of Semantic Web technologies is applied to allow for underspecified model transformation specifications as well as how the different assumptions on existing knowledge effect transformations. We demonstrate these aspects by a common case study.
Together with the latest efforts in publishing Linked (Open) Data, legal issues around publishing and consuming such data are gaining increased interest. Particular areas of interest include (i) how to define more expressive access policies which go beyond common licenses, (ii) how to introduce pricing models for online datasets (for non-open data) and (iii) how to realize (i)+(ii) while providing descriptions of respective meta data that is both human readable and machine processable. In this paper, we show based on different examples that the Open Digital Rights Language (ODRL) Ontology 2.0 is able to address all previous mentioned issues, i.e. is suitable to express a large variety of different access policies for Linked Data. By defining policies as ODRL in RDF we aim for (i) higher flexibility and simplicity in usage, (ii) machine/human readability and (iii) fine-grained policy expressions for Linked (Open) Data.
Large interconnected technical systems (e.g. railway networks, power grids, computer networks) are often configured with the help of multiple configurators, which store their configurations in separate databases based on heterogenous domain models (ontologies). When users want to ask queries over several distributed configurations, these domain models need to be aligned. To this end, standard mechanisms for ontology and data integration are required that enable combining query answering with reasoning about these distributed configurations. In this paper we describe our experience with using standard Semantic Web technologies (RDFS and SPARQL) in such a context.
The World Wide Web Consortium (W3C) as the main standardization body for Web standards has set a particular focus on publishing and integrating Open Data. In this article we will explain various standards from the W3C's Semantic Web activity and the - potential - role they play in the context of Open Data: RDF, as a standard data format for publishing and consuming structured information on the Web; the Linked Data principles for interlinking RDF data published across the Web and leveraging a Web of Data; RDFS & OWL to describe vocabularies used in RDF and for describing mappings between such vocabularies. We conclude with a review of current deployments of these standards on the Web, particularly within public Open Data initiatives, and discuss future potential, risks, and challenges.
Mix’n’Match is a framework to combine different ontology matchers in an iterative fashion for improved combined results: starting from an empty set of alignments, we aim at iteratively supporting in each round, matchers with the combined results of other matchers found in previous rounds, aggregating the results of a heterogeneous set of ontology matchers.
The existence of a standardized ontology alignment format promoted by the Ontology Alignment Evaluation Initiative (OAEI) potentially enables different ontology matchers to be combined and used together. Along these lines, we present a novel architecture for combining ontology matchers based on iterative calls of off-the-shelf matchers that exchange information in the form of reference mappings in this standard alignment format. However, we argue that only a few of the matchers contesting in the past years’ OAEI campaigns actually allow the provision of reference alignments to support the matching process. We bypass this lacking functionality by introducing an alternative approach for aligning results of different ontology matchers using simple URI replacement in the aligned ontologies. We experimentally prove that our iterative approach benefits from this emulation of reference alignments.