We generate the extended CSV file from the original input file by adding extra columns for the disambiguated entities. These extra columns are named *_url for disambiguations from primary knowledge base and *_alternative_urls for disambiguations from other knowledge bases (there can be more values separated by the space), where * stands for the name of the original column which is disambiguated. Together with Annotations (JSON) file it conforms to the CSV on the Web standard.

Example

	District;			City;		POP_FOR_NAT;	POP_TOTAL;	City_url;	District_url;	
	Graz-Stadt;			Graz;		46952;			269997;		...;		...;
	Deutschlandsberg;	Aibl;		39;				1386;		...;		...;

Export process

In the export proces all chosen cell annotations from the Result are read. Annotation for primary knowledge base is written to the *_url column (newly created, * stands for name of the corresponding original column), other annotations are written to the *_alternative_urls column. If the annotation for primary knowledge base is empty, one of other annotations is moved to the *_url column.

Statistical data specifics

When statistical data are processed, one more column (apart from columns with disambiguated URLs) is added to the extended CSV file. Its header name is "OBSERVATION_url" and values in cells are generated according to this template: "{kbInsertDataElementPrefix}observation/{UUID}", where kbInsertDataElementPrefix is fetched from the KnowledgeBase configuration (property called kb.insert.prefix.data.element) and UUID is generated randomly. These URLs then serve as subjects in triples describing particular observations (every row is one observation).