Load DBpedia locally using Jena TDB?

Question

I need to perform a query against DBpedia:

SELECT DISTINCT ?poi ?lat ?long ?photos ?template ?type ?label WHERE {
  ?poi  <http://www.w3.org/2000/01/rdf-schema#label> ?label .
  ?poi <http://www.w3.org/2003/01/geo/wgs84_pos#lat> ?lat .
  ?poi <http://www.w3.org/2003/01/geo/wgs84_pos#long> ?long .
  ?poi <http://dbpedia.org/property/hasPhotoCollection> ?photos .                      
  OPTIONAL {?poi <http://dbpedia.org/property/wikiPageUsesTemplate> ?template } .
  OPTIONAL {?poi <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> ?type } .
  FILTER ( ?lat > x && ?lat < y &&
           ?long > z && ?long < ω && 
           langMatches( lang(?label), "EN" ))
} 

I'm guessing this information is scattered among different dumps (.nt) files and somehow the SPARQL endpoint serves us with a result set. I need to download these different .nt files locally (not all DBpedia), perform only once my query and store the results locally (I don't want to use the SPARQL endpoint).

• What parts of Jena should I use for this one run?

I m a bit confused reading from this post:

So, you can load the entire DBPedia data into a single TDB location on disk (i.e. a single directory). This way, you can run SPARQL queries over it.

• How do I load the DBpedia into a single TDB location, in Jena terms, if we got three .nt DBpedia files? How do we apply the above query on those .nt files? (Any code would help.)

• Example, is this wrong?

 String tdbDirectory = "C:\\TDB";
 String dbdump1 = "C:\\Users\\dump1_en.nt";
 String dbdump2 = "C:\\Users\\dump2_en.nt";
 String dbdump3 = "C:\\Users\\dump3_en.nt";
 Dataset dataset = TDBFactory.createDataset(tdbDirectory);
 Model tdb = dataset.getDefaultModel(); //<-- What is the default model?Should I care?
 //Model tdb = TDBFactory.createModel(tdbdirectory) ;//<--is this prefered?
 FileManager.get().readModel( tdb, dbdump1, "N-TRIPLES" );
 FileManager.get().readModel( tdb, dbdump2, "N-TRIPLES" );
 FileManager.get().readModel( tdb, dbdump3, "N-TRIPLES" );
 String q = "my big fat query";
 Query query = QueryFactory.create(q);
        QueryExecution qexec = QueryExecutionFactory.create(query, tdb);
        ResultSet results = qexec.execSelect();
         while (results.hasNext()) {
         //do something significant with it
 }
qexec.close()
tdb.close() ;
dataset.close();

• In the above code we used "dataset.getDefaultModel" (to get the default graph as a Jena Model). Is this statement valid? Do we need to create a dataset to perform the query, or should we go with TDBFactory.createModel(tdbdirectory)?

Answer 1

To let Jena index locally :

/** The Constant tdbDirectory. */
public static final String tdbDirectory = "C:\\TDBLoadGeoCoordinatesAndLabels"; 

/** The Constant dbdump0. */
public static final String dbdump0 = "C:\\Users\\Public\\Documents\\TDB\\dbpedia_3.8\\dbpedia_3.8.owl";

/** The Constant dbdump1. */
public static final String dbdump1 = "C:\\Users\\Public\\Documents\\TDB\\geo_coordinates_en\\geo_coordinates_en.nt";

 ...

Model tdbModel = TDBFactory.createModel(tdbDirectory);<\n>

/*Incrementally read data to the Model, once per run , RAM > 6 GB*/
FileManager.get().readModel( tdbModel, dbdump0);
FileManager.get().readModel( tdbModel, dbdump1, "N-TRIPLES");
FileManager.get().readModel( tdbModel, dbdump2, "N-TRIPLES");
FileManager.get().readModel( tdbModel, dbdump3, "N-TRIPLES");
FileManager.get().readModel( tdbModel, dbdump4, "N-TRIPLES");
FileManager.get().readModel( tdbModel, dbdump5, "N-TRIPLES");
FileManager.get().readModel( tdbModel, dbdump6, "N-TRIPLES");
tdbModel.close();

To query Jena:

String queryStr = "dbpedia query ";

Dataset dataset = TDBFactory.createDataset(tdbDirectory);
Model tdb = dataset.getDefaultModel();

Query query = QueryFactory.create(queryStr);
QueryExecution qexec = QueryExecutionFactory.create(query, tdb);

/*Execute the Query*/
ResultSet results = qexec.execSelect();

while (results.hasNext()) {
    // Do something important
}

qexec.close();
tdb.close() ;