| Abstract |
Semantic Web data always commits to a particular view of the world. The same can be said about the mappings between various datasets: they are generated in the context of specific tasks. In the modelling and alignment process, decisions are made at different stages: (1) when data is converted to RDF, and (2) when alignments are generated. Unfortunately, these decisions are not made transparent. This limits users of Linked Data in their need to precisely select those links that fit their criteria. They simply do not have enough information to do so. We propose to model RDF-data in layers to make explicit the decisions made at each stage. These layers are: data, domain, correspondence and lens. Additionally, we propose a flexible content ontology design pattern for representing links between datasets that allows multiple alternative alignments to coexist. We use this rich representation to generate all potentially reasonable links between datasets, and then use task specific link selection criteria (expressed as simple SPARQL queries) to select the lens that contains the appropriate links for a particular task and context. We illustrate the validity of this approach in the field of Science, Technology and Innovation Studies. The goal of this research is not to develop another tool specialised in solving a particular mapping problem in a particular context. Instead, we aim to develop a framework that allows users to select context-sensitive alignments, because of the diversity of problems that need to be addressed in a variety of context-specific solutions. In other words, we aim at an approach that could cope with multiple alignment problems in a variety of contexts. |