2.5 URIGenerator

###Introduction The URI generation step transforms a pair of strings - extracted by a pair of learned XPATHs - into a triple given a predicate. Furthermore, it outputs a quadruple which is providing provenance information.

In this case, we implemented a URI generation based on AGDISTIS - Agnostic Named Entity Disambiguation. AGDISTIS is a state-of-the-art Named Entity Disambiguation tool which is based on string similarities, i.e., 3-gram similarity, breadt-first search in DBpedia and the HITS algorithm to rank candidate entities. More Information can be found in the paper.

Parameters

All URIDisambiguation instantiations have to follow the interface below. Especially, the classes have to implement the getTriples method with its parameters:

• Set result: A set of extraction results, i.e., XPATH pairs which contain a string for subject and a string for object.
• Property p: Predicate for the resulting triples.
• @return: This method returns a set of quadruples containing triples and with a provenance information.

public interface URIGenerator {
	
	Set<Quadruple<Node, Node, Node, String>> getTriples(Set<ExtractionResult> result, Property p) throws Exception;
}

Implementation

An implementation can be found in the URIGeneratorAGDISTIS.java.