Learning Ontologies from RDF Annotations Alexandre Delteil Catherine

Learning Ontologies from RDF Annotations Alexandre Delteil, Catherine Faron -Zucker, Rose Dieng ACACIA project, INRIA, 2004 Sophia Antipolis, France

TOC Introduction RDF & RDFS Background Ontology Example Approach to Ontology Learning Conclusion Future Work

Introduction Build ontologies from information extracted from RDF annotations “We have … a method to learn ontologies from RDF annotations by systematically generating the most specific generalization of all the possible sets of resources. ”

RDF Annotation Triplet statement (resource, property, value), (Njal, type, Cat) Easily represented as a graph XML syntax provided Property Resource or literal

XML Serialization of RDF Annotation Anonyms Resource

RDF Schema (RDFS) RDFS -> schema specification language Specifies ontological knowledge used in RDF statements Consists of a set of declarations of classes and properties Defines class and property hierarchies Multiple inheritance

RDFS Metamodel and RDFS Ontology

Pieces of Knowledge and Descriptions Piece of Knowledge -> set of nodes directly connected with the resource… Descriptionn -> largest set of nodes connected with the resource and having a path length <= n Complete Description -> the set of nodes connected to the resource through all possible properties

Piece of Knowledge Relative to Njal

D 1(Njal): Description

D 2(Njal): Description

Ontology Learning Systematically consider all concepts covering a set of resource nodes n RDF graph resource extraction techniques preliminary first step Group concepts and resources based on intensions and extensions Incrementally build generalization hierarchy

Building of S 1

D 1(Njal): Concept Hierarchy

Hierarch Based on Descriptions of Length N Construct triples of intensions and related extensions Iteratively join triple L 1 with triple in path Join all possible triples and paths Construct intensions of length n Build sets Sn from inclusion relations between node extensions

Building of S 2 from S 1

D 2(Njal): Concept Hierarchy

Conclusion Lacks clarity Gaps in logic in explanation, S 1 -> Ontology Relies on RDF annotations previously generated Result complexity can increase exponentially Requires no training data Little or no user input Implemented and tested inside European IST Comma Project

Future Work Inclusion of heuristics Insertion of domain specific criteria Graphical UI Bounding methods to reduce complexity RDF annotation generator