Statistical data often comes with inconsistencies: records of people with negative ages, pregnant males, and underaged car drivers often populate statistical databases. National Statistical Offices (NSO) encode knowledge to detect these inconsistencies in so-called edit rules. These days, there is an increasing number of statistical data being published and linked on the Web using the RDF Data Cube vocabulary. However, edit rules are hardly ever published together with data cubes on the Web. This causes two important problems: (a) quality of RDF Data Cube data cannot be assessed; and (b) reusability of edit rules is hampered. In this paper we present Linked Edit Rules (LER), a method that makes edit rules Web friendly and reusable as Linked Data. We show that LER can be easily linked, retrieved, reused, combined and executed to check quality and consistency of RDF Data Cubes, opening up the internal NSO validation processes to the Web.
|Titel||Proceedings of the 3rd International Workshop on Semantic Statistics co-located with 14th International Semantic Web Conference (ISWC 2015)|
|Redacteuren||Sarvan Capadisli, Franck Cotton, Armin Haller, Evangelos Kalampokis, Monica Scannapieco, Raphael Troncy|
|Status||Gepubliceerd - 2015|