KIT | KIT-Bibliothek | Impressum | Datenschutz

SOFYA: Semantic on-the-fly Relation Alignment

Koutraki, Maria; Preda, Nicoleta; Vodislav, Dan

Abstract:
Recent years have seen the rise of Web data, in particular Linked Data, with, up to now, more than 1000 datasets in the Linked Open Data Cloud (LOD). These datasets are mostly of entity-centric nature and are highly heterogeneous in terms of domains, language, schema, etc. Hence, the vision of uniformly querying such resources in the LOD has a long way to go. While equivalent entity instances across datasets are often linked by sameAs links, relations from different datasets and schemas are usually not aligned. In this paper, we propose an on-line instance-based relation alignment approach. The alignment may be performed during query execution and requires partial information from the datasets. We align relations to a target dataset using association rule mining approaches. We sample for equivalent entity instances with two main sampling strategies. Preliminary experiments, show that we are able to align relations with high accuracy, even if accessing the entire datasets is impossible or impractical.

Open Access Logo


Verlagsausgabe §
DOI: 10.5445/IR/1000092961
Veröffentlicht am 23.04.2019
Zugehörige Institution(en) am KIT Institut für Angewandte Informatik und Formale Beschreibungsverfahren (AIFB)
Publikationstyp Proceedingsbeitrag
Jahr 2016
Sprache Englisch
Identifikator ISBN: 978-389318070-7
ISSN: 2367-2005
KITopen-ID: 1000092961
Erschienen in 19th International Conference on Extending Database Technology, EDBT 2016; Bordeaux; France; 15 March 2016 through 18 March 2016. Ed.: I. Manolescu
Verlag OpenProceedings, Konstanz
Seiten 690-691
Serie Advances in Database Technology - EDBT ; 2016-March
Nachgewiesen in Scopus
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page