csv2rdf

csv2rdf is a Java based application, which relies on Apache Jena to convert tabular data to RDF. The binaries can be obtained by compiling the source code with MAVEN from command line, i.e.

Compiling

mvn clean install

Once the source code have been compiled a JAR named stlab.csv2rdf-1.0.jar is available in the target folder.

Usage

The JAR stlab.csv2rdf-1.0.jar can be used as a command line tool. The synopsis is the following

java -jar stlab.csv2rdf-1.0.jar [OPTIONS] CSV_FILE

The options available are the following:

-s,--separator <char>
The character used as separator within the CSV file (e.g. , or ;).
-m,--mapping <file>
A file providing the mapping between CSV columns and the properties of a target ontology/vocabulary.

Such a file must contain as set of key=value lines, where each key represents a column position in the source CSV (the counting of positions starts from index 1) and each value is a pair property-datatype composed of property URI form a target ontology or vocabulary and a datatype URI. The property-datatype pairs are separated by the character '>'. The datatype is optional, hence it is possible to provide the property URI only without any datatype. We remark that if no datatype is provided, then the tool tries to infer the more appropriate datatype for the value to transform to RDF. However, the datatype inference is performed for values that can be typed as integer, double or boolean only. The following is an example that associates the properties http://xmlns.com/foaf/0.1/givenName and http://dbpedia.org/ontology/birthDate to the second and third columns of a given source CSV. Additionally, the example specifies that the values associated with the property http://dbpedia.org/ontology/birthDate have to be typed as http://www.w3.org/2001/XMLSchema#date: java 2=http://xmlns.com/foaf/0.1/givenName 3=http://dbpedia.org/ontology/birthDate > http://www.w3.org/2001/XMLSchema#date

If no mapping or no value for a specific column in the mapping is provided then the URI for the predicate is generated from the column name by using either the default or the provided namespace.

-n,--namespace <uri>
The namespace to use for generating RDF objects. If no namespace is provided http://purl.org/example/ is used as default.
-o,--output <file>
The name of the file where to store the resulting RDF. If no file is provided the RDF is printed on screen.
-f, --format <string>
The format that the tool has to use in order to serialise the RDF output. Available aternatives are:
- TURTLE
- RDF/XML
- RDF/XML-ABBREV
- N-TRIPLES
- JSON-LD
- N3
- RDF/JSON

If no explicit format is provided, the output is serialised as TURTLE by default.

Example

Let's take the following table as a possible CSV file, named musicians.csv, to convert to RDF.

Name	Last name	Albums	Date of birth
Miles	Davis	102	1926-05-26
David	Bowie	90	1947-01-08
James	Brown	123	1933-05-03

Additionally, the following file, named mapping and containing key=property>datatype pairs, define the mapping to be used in order to generate the properties.

1=http://xmlns.com/foaf/0.1/givenName
2=http://xmlns.com/foaf/0.1/lastName
3=http://foo.org/myont/numberOfAlbums
4=http://dbpedia.org/ontology/birthDate > http://www.w3.org/2001/XMLSchema#date

Hence, the following line provides the example about how to use the tool from command line in order to obtain RDF from CSV and saving its content into a file named musicians.ttl. We suppose that the input CSV is actually a tab-separated file.

java -jar stlab.csv2rdf-1.0.jar -s '\t' -m mapping -o musicians.ttl musicians.csv

The execution of the tool with the arguments as provided in the example above produces the following RDF serialised by using the TURTLE syntax.

[ <http://dbpedia.org/ontology/birthDate>
         "1926-05-26"^^<http://www.w3.org/2001/XMLSchema#date> ;
 <http://foo.org/myont/numberOfAlbums>
         "102"^^<http://www.w3.org/2001/XMLSchema#int> ;
 <http://w3c/future-csv-vocab/row>
         1 ;
 <http://xmlns.com/foaf/0.1/givenName>
         "Miles" ;
 <http://xmlns.com/foaf/0.1/lastName>
         "Davis"
] .

[ <http://dbpedia.org/ontology/birthDate>
         "1947-01-8"^^<http://www.w3.org/2001/XMLSchema#date> ;
 <http://foo.org/myont/numberOfAlbums>
         "90"^^<http://www.w3.org/2001/XMLSchema#int> ;
 <http://w3c/future-csv-vocab/row>
         2 ;
 <http://xmlns.com/foaf/0.1/givenName>
         "David" ;
 <http://xmlns.com/foaf/0.1/lastName>
         "Bowie"
] .

[ <http://dbpedia.org/ontology/birthDate>
         "1933-05-03"^^<http://www.w3.org/2001/XMLSchema#date> ;
 <http://foo.org/myont/numberOfAlbums>
         "123"^^<http://www.w3.org/2001/XMLSchema#int> ;
 <http://w3c/future-csv-vocab/row>
         3 ;
 <http://xmlns.com/foaf/0.1/givenName>
         "James" ;
 <http://xmlns.com/foaf/0.1/lastName>
         "Brown"
] .

anuzzolese/csv2rdf

csv2rdf

Compiling

Usage

Example