Verifying the semantic coherence of the discovered alignment is a crucial task in ontology matching. Mapping selection is used at the end of the matching process in order to produce the final alignment. There are different strategies and methods for selecting mappings, that can be mainly classified into two categories. The first category is based on threshold filter and cardinality filter. The second category, called also semantic verification, uses semantic filter. It takes additional semantic information of entities in the input ontologies in consideration to select the best mappings. Verifying the semantic coherence of the discovered mappings is known as a crucial and challenging task namely in large scale ontology matching because almost all reasoning systems fail or cannot completely classify large ontologies. In this paper, we present our latest work in the field of semantic verification. In order to effectively detect explicit conflicts among a set of mappings, especially in the large scale ontology matching, we perform a structural indexing for the both to-be-matched ontologies. If disjoint relations are not found in those ontologies, we propose a semantically similarity measure to determine if two classes in a large ontology are potentially disjoint. Then, we define patterns to detect conflict mappings. Once the conflict set of mappings is located, an approximation algorithm is applied to remove this inconsistency. A prototype called YAM++ implementing these contributions has participated to OEAI2013 and has got top positions in all tracks where it has participated in. In this paper, we report and analyze the evaluation results on the effectiveness and efficiency of our approach for semantic verification in large scale OAEI Large Biomedical track.
[1]
Isabel F. Cruz,et al.
AgreementMakerLight results for OAEI 2013
,
2013,
OM.
[2]
Graeme Hirst,et al.
Evaluating WordNet-based Measures of Lexical Semantic Relatedness
,
2006,
CL.
[3]
Christian Meilicke,et al.
Alignment incoherence in ontology matching
,
2011
.
[4]
Erhard Rahm,et al.
GOMMA results for OAEI 2012
,
2012,
OM.
[5]
Alessandro Colantonio,et al.
Concise: Compressed 'n' Composable Integer Set
,
2010,
Inf. Process. Lett..
[6]
David W. Conrath,et al.
Semantic Similarity Based on Corpus Statistics and Lexical Taxonomy
,
1997,
ROCLING/IJCLCLP.
[7]
J. Euzenat,et al.
Ontology Matching
,
2007,
Springer Berlin Heidelberg.
[8]
Jérôme Euzenat,et al.
Ontology Matching: State of the Art and Future Challenges
,
2013,
IEEE Transactions on Knowledge and Data Engineering.
[9]
David Sánchez,et al.
Ontology-based information content computation
,
2011,
Knowl. Based Syst..
[10]
Jan Nößner,et al.
CODI: Combinatorial Optimization for Data Integration: results for OAEI 2011
,
2010,
OM.
[11]
Thomas R. Gruber,et al.
Toward principles for the design of ontologies used for knowledge sharing?
,
1995,
Int. J. Hum. Comput. Stud..
[12]
Arie Shoshani,et al.
Optimizing bitmap indices with efficient compression
,
2006,
TODS.
[13]
Zohra Bellahsene,et al.
Extended Tversky Similarity for Resolving Terminological Heterogeneities across Ontologies
,
2013,
OTM Conferences.