HAL - INRIA :: [inria-00596210, version 3] Ontology Alignment at the Instance and Schema Level

inria-00596210, version 3

Ontology Alignment at the Instance and Schema Level

Fabian Suchanek () ¹, Serge Abiteboul () ², Pierre Senellart ³

N° RT-0408 (2011)

Abstract: We present PARIS, an approach for the automatic alignment of ontologies. PARIS aligns not only instances, but also relations and classes. Alignments at the instance-level cross-fertilize with alignments at the schema-level. Thereby, our system provides a truly holistic solution to the problem of ontology alignment. The heart of the approach is probabilistic. This allows PARIS to run without any parameter tuning. We demonstrate the efficiency of the algorithm and its precision through extensive experiments. In particular, we obtain a precision of around 90% in experiments with two of the world's largest ontologies.

1: LEO (INRIA Saclay - Ile de France)
INRIA – CNRS : UMR8623 – Université Paris XI - Paris Sud
2: Dahu (INRIA Saclay - Ile de France)
INRIA – Ecole Nationale Supérieure de Cachan – CNRS : UMR8643
3: Institut Télécom - Télécom ParisTech
Télécom ParisTech

inria-00596210, version 3
http://hal.inria.fr/inria-00596210
oai:hal.inria.fr:inria-00596210
From: Fabian Suchanek <>
Submitted on: Thursday, 18 August 2011 14:47:08
Updated on: Thursday, 18 August 2011 15:00:38

See detailed view

Associated documents

PDF :

PS :

arXiv: 1105.5516

Export

Bibtex EndNote TEI RefWorks

%F inria-00596210, version 3
%0 Report
%U http://hal.inria.fr/inria-00596210
%2 INFO:INFO_AI
%T Ontology Alignment at the Instance and Schema Level
%A Suchanek, Fabian
%A Abiteboul, Serge
%A Senellart, Pierre
%+ LEO - INRIA Saclay - Ile de France , Dahu - INRIA Saclay - Ile de France , Institut Télécom - Télécom ParisTech
%X We present PARIS, an approach for the automatic alignment of ontologies. PARIS aligns not only instances, but also relations and classes. Alignments at the instance-level cross-fertilize with alignments at the schema-level. Thereby, our system provides a truly holistic solution to the problem of ontology alignment. The heart of the approach is probabilistic. This allows PARIS to run without any parameter tuning. We demonstrate the efficiency of the algorithm and its precision through extensive experiments. In particular, we obtain a precision of around 90% in experiments with two of the world's largest ontologies.
%2 Nous présentons PARIS, un système automatique d'alignement d'ontologies. PARIS réalise non seulement l'alignement d'instances de deux ontologies, mais aussi l'alignement des relations et des classes. L'alignement d'instances et de relations s'enrichissent mutuellement. Notre approche, qui fournit ainsi une solution holistique au problème d'alignement d'ontologies, repose sur un modèle probabiliste. De ce fait, notre algorithme ne nécessite ni paramètre arbitraire ni réglage manuel. Nous démontrons l'efficacité de PARIS à travers des expériences sur des ontologies diverses; nous obtenons des niveaux de précisions d'approximativement 90% pour l'alignement de deux des plus grandes ontologies publiquement disponibles.
%2 H.: Information Systems/H.1: MODELS AND PRINCIPLES/H.1.1: Systems and Information Theory/H.1.1.1: Information theory
%G English
%2 Technical Report
%P 26
%8 2011-08-18
%K Alignment; Ontology; Disambiguation; RDFS; Instances; Relations
%Z Technical Report at INRIA RT-0408
%Z RT-0408
%Z ERC Webdam
%2 14826