BUILDING TAILORED ONTOLOGIES FROM VERY LARGE KNOWLEDGE RESOURCES

Victoria Nebot, Rafael Berlanga

2009

Abstract

Nowadays very large domain knowledge resources are being developed in domains like Biomedicine. Users and applications can benefit enormously from these repositories in very different tasks, such as visualization, vocabulary homogenizing and classification. However, due to their large size and lack of formal semantics, they cannot be properly managed and exploited. Instead, it is necessary to provide small and useful logic-based ontologies from these large knowledge resource so that they become manageable and the user can take benefit from the semantics encoded. In this work we present a novel framework for efficiently indexing and generating ontologies according to the user requirements. Moreover, the generated ontologies are encoded using OWL logic-based axioms so that ontologies are provided with reasoning capabilities. Such a framework relies on an interval labeling scheme that efficiently manages the transitive relationships present in the domain knowledge resources. We have evaluated the proposed framework over the Unified Medical Language System (UMLS). Results show very good performance and scalability, demonstrating the applicability of the proposed framework in real scenarios.

Paper Citation

in Harvard Style

Nebot V. and Berlanga R. (2009). BUILDING TAILORED ONTOLOGIES FROM VERY LARGE KNOWLEDGE RESOURCES . In Proceedings of the 11th International Conference on Enterprise Information Systems - Volume 2: ICEIS, ISBN 978-989-8111-85-2, pages 144-151. DOI: 10.5220/0001984001440151

in Bibtex Style

@conference{iceis09,
author={Victoria Nebot and Rafael Berlanga},
title={BUILDING TAILORED ONTOLOGIES FROM VERY LARGE KNOWLEDGE RESOURCES},
booktitle={Proceedings of the 11th International Conference on Enterprise Information Systems - Volume 2: ICEIS,},
year={2009},
pages={144-151},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001984001440151},
isbn={978-989-8111-85-2},
}

in EndNote Style

TY - CONF
JO - Proceedings of the 11th International Conference on Enterprise Information Systems - Volume 2: ICEIS,
TI - BUILDING TAILORED ONTOLOGIES FROM VERY LARGE KNOWLEDGE RESOURCES
SN - 978-989-8111-85-2
AU - Nebot V.
AU - Berlanga R.
PY - 2009
SP - 144
EP - 151
DO - 10.5220/0001984001440151