Slicing Linked Data by Extracting Significant, Self-describing Subsets: The DBpedia Case
The Linked Data cloud is a huge set of data sources, the same objects being instances of many different ontologies and being described by different overlapping concepts and properties. This paper shows an innovative approach to represent in an unambiguous and homogeneous way instances of such large, highly unpredictable, linked ontologies. Instead of making use of external, static and ad-hoc devised knowledge bases, the approach followed in this paper leverages an existing ontology in the Linked Data cloud to univocally identify all the instances of all linked ontologies. Following this kind of approach, different views over the same set of instances can be devised, depending of which spot of the cloud we choose to see things through.
KeywordsLinkedData Ontology SKOS
- 1.Calvanese, D., De Giacomo, G., Lembo, D., Lenzerini, M., Rosati, R.: Data Complexity of Query Answering in Description Logics (2006)Google Scholar
- 2.Perez, J., Arenas, M., Gutierrez, C.: Semantics and Complexity of SPARQL (2006)Google Scholar
- 3.Dean, J., Ghemawat, S.: MapReduce: Simplied Data Processing on Large Clusters (2004)Google Scholar