CODEWITHSUNDEEP
sparqljenaontologydbpedia
Elmira Khodaee
Elmira Khodaee

Reputation: 93

How to run a sparql when information are in different ModelRDB?

I am working in DBPedia. i want to run some sparql. so i downloaded files from here: http://wiki.dbpedia.org/services-resources/ontology.

As you see, DBpedia seperated its triples into different files. A file for rdf:type properties and another file for others properties.

I import each file into MySQL (Persist a model); Condsider i named models "A" and "B" in order! the code is below:

for DBpedia Ontology RDF type statements

 private Model AModel;
 private Model BModel;

 DBConnection connection = new DBConnection(DB_URL, DB_USER, DB_PASSWORD, DB_TYPE);
 AModel = maker.createModel("A",true);
 model.begin();
 InputStream in =this.getClass().getClassLoader().getResourceAsStream("some address /instance_types_en.nt);
 model.read(in,null);// Commit the database transactionmodel.commit();

------------

and again for DBpedia Ontology other A-Box properties

DBConnection connection = new DBConnection(DB_URL, DB_USER, DB_PASSWORD, DB_TYPE);
BModel = maker.createModel("B",true);
model.begin();
InputStream in =this.getClass().getClassLoader().getResourceAsStream("some address /mappingbased_properties_en.nt);
model.read(in,null);// Commit the database transactionmodel.commit();

so till here, we have two models, ones for DBpedia Ontology RDF type statements and another for DBpedia Ontology other A-Box properties.

The problen is where we have a sparql that need both inforamtion. (rdf:type properties and non rdf:type prpperties); such as:

Who was the wife of President Lincoln?

its equivalence sparql is:

PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX foaf: <http://xmlns.com/foaf/0.1/>
PREFIX onto: <http://dbpedia.org/ontology/>
SELECT ?uri ?string
WHERE 
{
    ?person rdf:type onto:President .
    ?person foaf:surname "Lincoln"@en .
    ?person onto:spouse ?uri.
        OPTIONAL {?uri rdfs:label ?string .}
        FILTER (lang(?string) = "en") 
}

so i should run this sparql in models. but jena just let me to perform sparql in just one model:

so i have to run it like:

QueryExecution qe1 = QueryExecutionFactory.create(query, AModel));

or 

QueryExecution qe1 = QueryExecutionFactory.create(query, BModel));

**

?person onto:spouse ?uri. is in AModel and ?person rdf:type onto:Presiden is in BModel!

So this query returns no record!

so what is the solution. my program generate the right sparql but i do not have the appropriate data set to return the response! Because i have different models, so i can not use Union operator here!!

Upvotes: 0

Views: 76

Answers (1)

AndyS
AndyS

Reputation: 16630

There are several approaches:

  1. Read all the data into the same model
  2. Use a union model (ModelFactory.createUnion or MultiUnion).
  3. Read the models separately into a RDF dataset (not possible with RDB).
  4. Create a Dataset which has both models separately in it.

If you are using a version of Jena which has DBConnection, some of these may not work as the code is so old.

Upvotes: 0

Related Questions