Paper Info
Title
Neighbour based structural proximity measures for ontology matching systems
Abstract
The evolution of semantic web leads to more heterogeneity and interoperability problems among the semantic data represented by ontologies. Ontology matching techniques play an important role in semantic web. The ontology matching techniques are used to handle heterogeneity among ontologies and to establish interoperability. The effectiveness of the ontology matching systems is evaluated by the match quality measured in terms of precision and recall. The efficiency of the ontology matching systems depends on the size of the input ontologies (number of concepts in ontology). The size of the ontologies being matched influences the efficiency in terms of execution time and may lead to out of memory error. Hence improving the efficiency of ontology matching system insists on reducing the concept match space which leads to less execution time. The concept match space could be reduced by decomposing the ontologies into disjoint clusters. In this paper, two new neighbour based structural proximity measures TNSP (Tversky based Neighbour Structural Proximity) and DNSP (Dice based Neighbour Structural Proximity) are proposed to form disjoint clusters of the ontology. The proposed measures reduce the number of computations required to identify structurally similar inter ontology concepts thereby improving the efficiency. This reduction in computation is achieved as each concept is compared only with neighbour concepts. The best neighbour combination for the proposed measures TNSP and DNSP is experimentally determined. The proposed measures were evaluated experimentally on real world large ontologies (mouse and human anatomy). The experiments prove that the proposed neighbour based structural similarity measures are more efficient than the existing structural similarity measures without compromising on effectiveness.
Year
DOI
Venue
2012
10.1145/2345396.2345569
ICACCI
Keywords
Field
DocType
ontology matching technique,large ontology,ontology matching system,execution time,structural proximity measure,proposed measure,input ontology,disjoint cluster,neighbour structural proximity,concept match space,semantic web,ontology,ontology matching,structural similarity
Ontology (information science),Ontology alignment,Ontology-based data integration,Data mining,Ontology,Interoperability,Computer science,Precision and recall,Semantic Web,Semantic data model
Conference
Citations 
PageRank 
References 
1
0.45
14
Authors
3
Name
Order
Citations
PageRank
K. Saruladha121.81
G. Aghila21911.85
B. Sathiya310.79