Hi Amany,

1) The DBpedia ontology linked from the latest download page (http://wiki.dbpedia.org/downloads-2016-10) includes labels in a number of languages. Each label is an rdfs:label. If you want to translate the ontology into another language, you should add new rdfs:label statements for each resource you wish to translate. That is, make a new label for a resource in your language, but please do not translate the English terms used in the resource URLs. Changing the resource URLs would change the dataset and stop you from having automatic linkages to other data that uses those URLs.

2) I typically download the parts of DBpedia that I need for a particular project, then make new data as necessary to link other data sets to the DBpedia resources. My new data always includes some form of schema (in RDFS) or ontology (in OWL) to describe it. Is that what you were asking?

3) I generally start by looking through the various files in the DBpedia downloads and map which relationships are in each file. I have made various human-readable graphs to show the relationships. I find having a hand-drawn graph of the relationships including the files they came from to be very helpful later when I'm writing SPARQL queries against the data. I often put the contents of each downloaded file into its own named graph to keep all the data straight. Doing that also helps you make readable SPARQL, even though scattering the data in different named graphs does make the queries longer.

Similarly, any data I make myself and any data from other data sets are in their own named graphs.

I hope that helps!

Regards,

Dave

Hello Dave

1- how can I know the achievement in specific version of DBpedia as Arabic or other language?

2- when you start work in your version of DBpedia, are you build your ontology from infoboxes or what? or directly using RDF files better than using ontology?

3- are you recommend codes of extraction in github

Best Regards