Structured knowledge creation for Urdu language: A DBpedia approach

Abstract

Wikipedia information is extracted by DBpedia and linked to other web resources as Linked Open Data, which is an important contribution to the field of semantics. As part of its internationalisation endeavour, DBpedia now has 20 language chapters that have been mapped to it; nonetheless, there have been very few attempts from Urdu. This article outlines the procedures and highlights the efforts put forward as the first contribution to the manual creation of Urdu mappings with DBpedia Ontology classes. Our approach led to an increase in the number of mapped infoboxes, thus enhancing the DBpedia. The mapping procedure is broken down into two parts. The infobox template is first mapped to the DBpedia ontology's relevant class, and then the attributes of the infobox are mapped to the properties of that class. In addition, alongside other mapped languages, Urdu labels are included to the description of Ontology classes. We have covered around a thousand properties and attributes of Urdu with English DBpedia Ontology on DBpedia mapping server.