Interlinking of Resources with Semantics.
Michael Hausenblas, Wolfgang Halb · 2008
Abstract. Semantic Web applications that are after utilising linkeddata datasets are often limited by the interlinking methodology available. Based on our experiences with building a linked-data dataset for humans and machines, we propose a new way of creating semantic links, labelled as User Contributed Interlinking. 1 Linking Open Data 101 Links are one of the key factors that made the Web of Documents successful. The same holds for the Web of Data; rather than having untyped hyperlinks between documents, RDF properties (such as rdfs:seeAlso) are interpreted as “semantic hyperlinks”. The linked data principles 1 basically capture this idea. The W3C Linking Open Data (LOD) project is a collaborative effort to bootstrap the building of linked-data datasets. In [1] the publication of linked data is described; further, the two basic approaches for creating links to other datasets are outlined: RDF links can either be set manually or generated by automated linking algorithms for large datasets. 2 User Contributed Interlinking With the User Contributed Interlinking (UCI) [2] we have proposed a novel approach for creating high-quality interlinks by relying on end-users. For large datasets, such as riese 2 where the entire European statistics are brought to the Semantic Web, it might appear impractical at first sight to manually generate interlinks to other datasets. However, we strongly believe that by applying the Wiki-principle end-users can be encouraged to contribute to linked datasets with similar enthusiasm as for example the case with Wikipedia. We have implemented a generalised UCI in a demonstrator called “interlinking of resources with semantics ” (irs), available at