Home // CONTENT 2014, The Sixth International Conference on Creative Content Technologies // View article

Refined Ontology Matching Methods for Special Data Integration

Authors:
Dénes Paczolay
András Bánhalmi
Ádám Zoltán Végh
Gábor Antal
Vilmos Bilicki

Keywords: semantic knowledge representation; local ontology generation; ontology matching; ontology alignment; ontology; Java to ontology; similarity measure

Abstract:
Ontology matching is an important area of research, since it has many applications like semantic webs, information extraction, data mining and reasoning. In most cases, the matching is done between thesauri of hierarchical concepts made for similar domains by different groups. A methodically similar, but technically different task is when ontology matching is used for system integration applied to generated ontologies. For data integration, we created a framework that generates ontologies got from a database schema or from the source code itself, these being called local and global ontologies. To complete the data integration process, only ontology matching has to be done semi-automatically; all the other tasks are carried out automatically. However, the generated ontologies have some special features, namely mixed languages in the name, special abbreviations, and special structures in the names generated. For ontologies like these, the common matching methods which have the best performance on average when these are applied to other tasks, perform much worse in the case of integration. In order to improve the accuracy, we propose novel similarity measuring and ontology matching methods.

Pages: 53 to 58

Copyright: Copyright (c) IARIA, 2014

Publication date: May 25, 2014

Published in: conference

ISSN: 2308-4162

ISBN: 978-1-61208-342-1

Location: Venice, Italy

Dates: from May 25, 2014 to May 29, 2014